List of Archived Posts

2010 Newsgroup Postings (01/11 - 01/24)

How long before Microsoft goes the way of DEC (and in part, IBM)?
Korean bank Moves back to Mainframes (...no, not back)
Korean bank Moves back to Mainframes (...no, not back)
Korean bank Moves back to Mainframes (...no, not back)
Happy DEC-10 Day
Happy DEC-10 Day
Bookshelves under BookMangler
Happy DEC-10 Day
Happy DEC-10 Day
Happy DEC-10 Day
Happy DEC-10 Day
Happy DEC-10 Day
Korean bank Moves back to Mainframes (...no, not back)
Larrabee delayed: anyone know what's happening?
360 programs on a z/10
security and online banking
How long for IBM System/360 architecture and its descendants?
How long for IBM System/360 architecture and its descendants?
security and online banking
STEM crisis
Happy DEC-10 Day
security and online banking
How long for IBM System/360 architecture and its descendants?
Happy DEC-10 Day
Happy DEC-10 Day
Happy DEC-10 Day
Happy DEC-10 Day
Happy DEC-10 Day
Happy DEC-10 Day
Happy DEC-10 Day
Happy DEC-10 Day
Happy DEC-10 Day
Happy DEC-10 Day
Happy DEC-10 Day
Happy DEC-10 Day
Happy DEC-10 Day
Happy DEC-10 Day
Happy DEC-10 Day
Happy DEC-10 Day
Happy DEC-10 Day
Daylight Savings Time again
Happy DEC-10 Day
Happy DEC-10 Day
Daylight Savings Time again
sysout using machine control instead of ANSI control
sysout using machine control instead of ANSI control
Oldest Instruction Set still in daily use?
Daylight Savings Time again
Happy DEC-10 Day
search engine history, was Happy DEC-10 Day
Happy DEC-10 Day
Source code for s/360
The Shannon limit
Happy DEC-10 Day
ISPs could cut spam easily, says expert
Oldest Instruction Set still in daily use?
Happy DEC-10 Day
Happy DEC-10 Day
How long for IBM System/360 architecture and its descendants?
Happy DEC-10 Day
Happy DEC-10 Day
Source code for s/360 [PUBLIC]
Happy DEC-10 Day
Source code for s/360 [PUBLIC]
SYSENTER/SYSEXIT_vs._SYSCALL/SYSRET
Source code for s/360 [PUBLIC]
SYSENTER/SYSEXIT_vs._SYSCALL/SYSRET
How long for IBM System/360 architecture and its descendants?
Happy DEC-10 Day
Happy DEC-10 Day
Happy DEC-10 Day
Happy DEC-10 Day
Happy DEC-10 Day
Happy DEC-10 Day
Happy DEC-10 Day
Windows plagued by 17-year-old privilege escalation bug
Happy DEC-10 Day
search engine history, was Happy DEC-10 Day
Happy DEC-10 Day
Happy DEC-10 Day
Happy DEC-10 Day
Happy DEC-10 Day
Oldest Instruction Set still in daily use?
Happy DEC-10 Day
Happy DEC-10 Day
search engine history, was Happy DEC-10 Day
Oldest Instruction Set still in daily use?
"The Naked Mainframe" (Forbes Security Article)
"The Naked Mainframe" (Forbes Security Article)
Remember Ed Curry!
Happy DEC-10 Day
Happy DEC-10 Day
Who's to Blame for the Meltdown?
search engine history, was Happy DEC-10 Day
Oldest Instruction Set still in daily use?
Oldest Instruction Set still in daily use?
"The Naked Mainframe" (Forbes Security Article)
"The Naked Mainframe" (Forbes Security Article)
"The Naked Mainframe" (Forbes Security Article)
"The Naked Mainframe" (Forbes Security Article)
"The Naked Mainframe" (Forbes Security Article)

How long before Microsoft goes the way of DEC (and in part, IBM)?

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: How long before Microsoft goes the way of DEC (and in part, IBM)?
Newsgroups: alt.folklore.computers
Date: Mon, 11 Jan 2010 21:41:55 -0500
Michael Wojcik <mwojcik@newsguy.com> writes:
A joint venture from IBM, SCO, Sequent, and others to create a single flavor of Unix for a variety of platforms. The impression in the late 1990s was that Windows was doing well on smaller machines partly because the Unix market was too fragmented.

OSF wiki page:
http://en.wikipedia.org/wiki/Open_Software_Foundation

from above:
The founding of the organization was largely seen as a response to the collaboration between AT&T and Sun Microsystems on UNIX System V Release 4, and a fear that other vendors would be locked out of the standardization process. This led Scott McNealy of Sun to quip that "OSF" really stood for "Oppose Sun Forever." The competition between the opposing versions of UNIX systems became known as the UNIX wars.

... snip ...

Common Open Software Environment
http://en.wikipedia.org/wiki/Common_Open_Software_Environment

from above:
By the early 1990s, the major Unix players had begun to realize that the standards rivalries known as the Unix wars were causing all participants more harm than good, leaving Unix open to emerging competition from Microsoft. The COSE initiative in 1993 can be considered to be the first unification step and the merger of the Open Software Foundation (OSF) and X/Open in 1996 as the ultimate step in the end of those skirmishes.

... snip ...

Open Group wiki page:
http://en.wikipedia.org/wiki/The_Open_Group

from above:
It was formed when X/Open merged with the Open Software Foundation in 1996. The Open Group is most famous as the certifying body for the UNIX trademark, in the past the group was best known for its publication of the Single UNIX Specification paper, which extends the POSIX standards and is the official definition of UNIX.

... snip ...

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Korean bank Moves back to Mainframes (...no, not back)

Refed: **, - **, - **, - **, - **
From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Korean bank Moves back to Mainframes (...no, not back)
Newsgroups: bit.listserv.ibm-main
Date: Tue, 12 Jan 2010 10:21:29 -0500
sam@PSCSI.NET (Sam Siegel) writes:
Every state has laws regarding the retention of data related to the conduct of business. The amount of time is typically 3 to 7 years. No keeping the receipts (or copies thereof) could create legal problems as well.

re:
http://www.garlic.com/~lynn/2010.html#98 Korean bank Moves back to Mainframes (...no, not back)

a lot of record retention is by UCC which most states follow ... aka like for checks:
http://www.bankersonline.com/compliance/gurus_cmp1001l.html

above references
if the items are not returned to customer

... in the credit card slip case ... both the consumer and the merchant have copies.

the electronic record of the transaction data is kept (by the issuing bank) ... question of what wasn't kept was the merchant's paper slip copy with signature &/or electronic image of same.

the issue was resolving (potentially legal) disputes ... what side has burden of proof and what kind of proof. merchant not having the signed slip effectively resolves on behalf of the consumer (having the signed slip doesn't mean that it resolves on behalf of the merchant ... the merchant still has to show that it is the consumer's signature).

other items are like how long does consumer have to dispute items.

in any case, standard "reg. E" places burden of proof on merchant

one of the interesting flyers in the 90s was proposal about digitally signed, public key transactions for internet transactions. consumers would pay $100/annum for their digital certificate ... and in effort to sweeten the deal for merchants to install the technology ... the burden of proof (in disputes) for public key transactions ... would be switched from merchant to consumer. the question was raised ... why would the consumer pay $100/annum for something that would switch the burden of proof to them.

there has been some amount of churn in the UK with their chip payment card about something analogous ... where the dispute burden of proof is now effectively on the consumer.

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Korean bank Moves back to Mainframes (...no, not back)

From: lynn@GARLIC.COM (Anne & Lynn Wheeler)
Subject: Re: Korean bank Moves back to Mainframes (...no, not back)
Newsgroups: bit.listserv.ibm-main
Date: 12 Jan 2010 07:48:51 -0800
lynn@GARLIC.COM (Anne & Lynn Wheeler) writes:
however, by at least the early 90s, there were cases of compromised end-points recording valid information (done during the process of valid transactions). these operations tended to be more large scale wholesale operations ... getting information for tens of thousand (or millions) ... rather than a few tens.

re:
http://www.garlic.com/~lynn/2010.html#97 Korean bank Moves back to Mainframes (...no, not back)

skimming news item from today:

ATM Skimming Incidents Increase
http://www.bankinfosecurity.com/articles.php?art_id=2059

frequently these are external attachments specifically targeting magstripe ... however, there have been lots of cases where collecting technology has been installed inside the end-point (pos terminal or atm cash machine). cases have included modification of machines already installed, replacing machine with modified machine, installing modification at time of manufacture ... or even criminal front organization manufacturing machines and selling them on open market (or on gray market ... copy of some other vendors machine).

criminal front manufacturers have even sold such machines "at cost" (undercutting competition) because they are planning on making up the profit with fraudulent transactions.

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Korean bank Moves back to Mainframes (...no, not back)

Refed: **, - **, - **, - **
From: lynn@GARLIC.COM (Anne & Lynn Wheeler)
Subject: Re: Korean bank Moves back to Mainframes (...no, not back)
Newsgroups: bit.listserv.ibm-main
Date: 12 Jan 2010 08:26:53 -0800
lynn@GARLIC.COM (Anne & Lynn Wheeler) writes:
there has been some amount of churn in the UK with their chip payment card about something analogous ... where the dispute burden of proof is now effectively on the consumer.

re:
http://www.garlic.com/~lynn/2010b.html#1 Korean bank Moves back to Mainframes (...no, not back)
http://www.garlic.com/~lynn/2010b.html#2 Korean bank Moves back to Mainframes (...no, not back)

there was recent case in the UK where an individual needed a copy of the ATM machine video recording to prove that they didn't make the withdrawal ... since the bank wasn't able to find the recording ... it was decided in favor of the bank (and against the individual).

there have been comments that care taken regarding video recording might be significantly different if the bank was required to show the video recording to prove it was the individual (as opposed to the individual getting a copy from the bank to prove it wasn't them).

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Happy DEC-10 Day

Refed: **, - **, - **, - **, - **, - **, - **, - **, - **, - **
From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Happy DEC-10 Day
Newsgroups: alt.sys.pdp10, alt.folklore.computers
Date: Tue, 12 Jan 2010 15:40:09 -0500
Pat Farrell <pfarrell@pfarrell.com> writes:
Isn't this just 20-20 hindsight? DEC was heavily into the whole OSI/ISO process, driving parts of the spec. It was not clear in the early 80s whether TCP/IP or OSI/ISO would actually scale to world wide.

The US Government mandated OSI/ISO in about 85.

Both the Government and DEC were shown to be wrong, but I don't agree that it was obvious at the time.


I would claim it was very obvious ... OSI/ISO was "old technology", high error copper wire and private networks ... somewhat 60s&70s arpanet ... one of the things learned from arpanet was the need for internetworking ... which OSI/ISO didn't have.

interop '88 was supposedly at least venue for demonstrating ip interoperability ... but there were lots of vendors with booths showing off bits and pieces of OSI/ISO technology (possibly because of the federal gov. gosip mandates) ... misc. past posts mentioning
http://www.garlic.com/~lynn/subnetwork.html#interop88

i had some tcp/ip stuff in booth at interop '88 ... but it was a different corporation (than the one that was paying me).

another scenario from the mid-80s was frequent comments about ISO not requiring demonstration of actual implementation for standard (this was with regard to observations by people attempting full implementation was nearly impossible as well as impractical ... resulting in lots of people then doing apoligies that it was purely a model and nobody was expected to actually do an implementation). By comparison, it was pointed at that IETF requires interoperable implementations for proceeding to standard.

the other comment about gov, DEC, as well as certain sectors of IBM (as well as others) ... was somewhat similar to recent observation about people at rarified levels believing if they say it is so ... then it will be so ... recent reference on who to blame for IBM's asciii v. ebcdic situation, referenced in this post
http://www.garlic.com/~lynn/2009s.html#63 CAPS Fantasia
and:
EBCDIC and the P-BIT (The Biggest Computer Goof Ever)
http://www.bobbemer.com/P-BIT.HTM

I was on the XTP technical advisory board ... and there was an attempt to take XTP to x3s3.3 (iso chartered us body responsible for standards corresponding to OSI level 3&4) as HSP (high-speed protocol). x3s3.3 was under ISO mandate that no standards could be done for anything that didn't correspond to OSI. HSP was rejected because if failed to correspond to OSI for at least:

1) it supported internetworking ... something that doesn't exist in OSI
2) it supported LAN MAC interface ... something that doesn't exist in OSI ... sits approx. in the middle of OSI level 3
3) it went directly from level four to LAN MAC interface, bypassing OSI level 3/4 interface

misc. past posts mentioning OSI, HSP, and/or XTP
http://www.garlic.com/~lynn/subnetwork.html#xtphsp

in fact, the corporate communication group attempted to make sure there was no participation in XTP/HSP at all ... to say nothing of my position on the XTP TAB ... old email communication group position on HSP
http://www.garlic.com/~lynn/2006i.html#email890901
in this post
http://www.garlic.com/~lynn/2006i.html#17 blast from the past on reliable communication

other posts in the above thread
http://www.garlic.com/~lynn/2006h.html#46 blast from the past, tcp/ip, project athena and kerberos
http://www.garlic.com/~lynn/2006i.html#16 blast from the past, reliable communication
http://www.garlic.com/~lynn/2006i.html#18 blast from the past on reliable communication
http://www.garlic.com/~lynn/2006i.html#19 blast from the past on reliable communication
http://www.garlic.com/~lynn/2006i.html#20 blast from the past on reliable communication
http://www.garlic.com/~lynn/2006i.html#21 blast from the past on reliable communication

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Happy DEC-10 Day

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Happy DEC-10 Day
Newsgroups: alt.sys.pdp10, alt.folklore.computers
Date: Tue, 12 Jan 2010 16:05:56 -0500
Eric Chomko <pne.chomko@comcast.net> writes:
Why didn't Alta Vista become Google before Google?

it did ... the issue was possibly that DEC influence bogged it down ... while Google eventually came along w/o that baggage?

a couple items from long ago and far away

Date: Tue, 08 Sep 1998 01:16:07 -0700
To: google@palo.Stanford.edu
From: "L a r r y . P a g e" <page@cs.stanford.edu>
Subject: Google Downtime
Cc: google@palo.Stanford.edu

Thanks very much for letting us know Google was down. The system is now back up.

The whole Google team was on vacation at the Burning Man festival for Labor Day weekend. Unfortunately, we were out of email, web, and cell phone contact due to the remote location. We were hoping the system would stay up, but what seems to have been a loose disk cable caused our system to go down. We plan to add additional staff and redundant machines to improve the reliability in the future.

Also, we plan to replace the current index with a new one hopefully this week.

Thanks for using Google, and let us know if you have any other comments.

Regards, -Larry


... snip ... top of post, old email index.

and more than two years earlier ...

Date: Sat, 23 Mar 96 11:34:16 -0500
Subject: Internet Search and Indexing

BEHIND THE SCREENS- THE ALTA VISTA INTERNET SEARCH ENGINE

Like all the world's pivotal innovations, Alta Vista started life on the back of a napkin. Just about a year ago, Louis Monier and Paul Flaherty, both engineers at Digital's Palo Alto research labs, sat down to lunch and got talking about big numbers. The newspapers were full of Internet stories at the time, and editorials were predicting that the amount of information online would soon be too much to imagine, much less quantify. Meanwhile, Digital had just launched the Turbo Laser, an Alpha-based server with a 64-bit address bus - theoretically capable of addressing 17 billion gigabytes.

"We just had this crazy idea," recalls Monier, "of putting two and two together." Twelve months have since passed and the idea - crazy or otherwise - has become a reality. Alta Vista is now online at
http://altavista.digital.com. Its Turbo Laser, an eight-processor AlphaServer 8400 5/300 with a massive 6GB memory and 210GB of RAID, provides the largest full-text searchable index currently available on the Web.

Scooter the spider

Monier, now principal engineer on the project, spent much of last summer fashioning a web crawler capable of retrieving the contents of the entire Net. His scratch-built design, called Scooter, is a multi-threaded spider capable of retrieving as many as 1,000 documents simultaneously. It runs from a single Alpha workstation (a DEC 3000 Model 900 with 1GB of memory and 30GB of RAID) at Palo Alto, and has been designed to be a good web citizen - it obeys the Standard for Robot Exclusion (see D3 p39, last month) and avoids hitting the same site repeatedly.

Prototype index

While work on Scooter was under way, Monier hooked up with Mike Burrows, a fellow researcher at Palo Alto. Burrows had developed a prototype indexing technology as part of another project, and this proved crucial. Monier describes it as, "Probably the fastest and best indexing technology in the world." The indexer, which runs on the Turbo Laser, can handle about 1GB of text per hour, building a database that preserves the full text of the pages it has read. This is the main bottleneck of the whole process, and Scooter could actually run much faster. The indexer has so far processed around 100GB - retrieved from around 22 million pages of text. The resulting index is around 33GB. "The fact that we provide a full-text search is the biggest factor in keeping it so big," Monier says. A full-text index allows a number of techniques not possible by other means, such as


... snip ... top of post, old email index.

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Bookshelves under BookMangler

Refed: **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **
From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Bookshelves under BookMangler
Newsgroups: bit.listserv.ibm-main
Date: Tue, 12 Jan 2010 19:01:56 -0500
edjaffe@PHOENIXSOFTWARE.COM (Edward Jaffe) writes:
The story, as told by John Ehrman, is that the POO got so big, it broke the book build software and nobody at IBM has the time, inclination, or knowledge to fix it. :-(

it would be fun to get a look at it to fix generation of html ... POO has been subset of the architecture book ... which has been twice as large ... started out as cms script file with conditionals ... that command line arguments to the cms script command would either format the full document or just the POO subset.

last spring I had done a lot with the transcripts of the pecora hearings (senate banking hearings in the wake of '29 crash ... leading up to Glass-Steagall) ... with a whole lot of cross-indexing and generated loads of hrefs. the original scanned transcripts were six volumes with 2345 pgs total and 20 volumes with 9296 pgs total.

the original document wasn't the best ... so the scan wasn't outstanding and several places the OCR of the scanned pages is very low quality ... so the individual HTML'ed pages from the OCR, periodically have a lot of garbage; as a result I put in each HTML'ed page a HREF reference back to the corresponding page in the scan'ed document (whole thing is under two gbytes, most of which are the original scanned files).

by comparison, Z -07 POO PDF file says 1344 pages ... for the heck of it I just started a "save as text" ... which is going quite slow ... a lot of the formating & figures are lost in "save as text"

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Happy DEC-10 Day

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Happy DEC-10 Day
Newsgroups: alt.sys.pdp10, alt.folklore.computers
Date: Tue, 12 Jan 2010 19:29:26 -0500
glen herrmannsfeldt <gah@ugcs.caltech.edu> writes:
There was a manual on the PDP-10 hardware (instruction formats, opcodes, etc.) It was way too expensive, though, for me to buy. IBM kept their Principles of Operation manuals pretty reasonably priced until very recently.

but softcopy is readily available ... realtime thread in ibm-main about latest softcopy is pdf only ... online HTML seems to be discontinued
http://www.garlic.com/~lynn/2010b.html#6 Bookshelves under BookMangler

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Happy DEC-10 Day

Refed: **, - **, - **, - **, - **, - **, - **, - **
From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Happy DEC-10 Day
Newsgroups: alt.sys.pdp10, alt.folklore.computers
Date: Tue, 12 Jan 2010 21:39:55 -0500
Mark Crispin <mrc@panda.com> writes:
When people ask, "what were they thinking" when they look at the email system in use at the White House, it comes back to GOSIP. Both Clinton and Bush inherited a true nightmare which political detractors have attempted to present as evidence of conspiracy when it was really the lingering effects of GOSIP.

re:
http://www.garlic.com/~lynn/2010b.html#4 Happy DEC-10 Day

but in early/mid 80s, executive branch and several other organizations was PROFS. there is folklore that part of ollie's problem was the extensive profs backup procedures (actually general backup, but what the heck).

there were some rather derogatory labels applied to old-time telco mentality that were attributed to have been heavily involved with OSI/ISO standard. one of the more polite term referenced in this older post
http://www.garlic.com/~lynn/2003j.html#73 1950s AT&T/IBM lack of collaboration?

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Happy DEC-10 Day

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Happy DEC-10 Day
Newsgroups: alt.sys.pdp10, alt.folklore.computers
Date: Tue, 12 Jan 2010 22:26:21 -0500
glen herrmannsfeldt <gah@ugcs.caltech.edu> writes:
Though there are stories about features that were removed from S/370 because they were too hard to implement on some models.

i periodically rail about dropping r/o segment protect and various other features from 370 ... when 370/165 engineers ran into various problems and delays retroffiting 370 virtual memory (field install/upgrade from 165 to 165II)

(vm370)cms had already been reorged to utilze r/o segment protect and various other "new" features ... when it was decided to drop the features to gain back six months in 370 virtual memory (& 165) announce schedule

part of this dates back to doing the modifications in cp67 to provide 370 virtual machines including full virtual memory architecture ... this was running and in regular use a year before the first engineering 370 (145) virtual memory hardware was operational.

cp67 virtual 370 was also early distributed project between endicott and cambridge ... using (internal) network connections.

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Happy DEC-10 Day

Refed: **, - **, - **, - **
From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Happy DEC-10 Day
Newsgroups: alt.sys.pdp10, alt.folklore.computers
Date: Tue, 12 Jan 2010 23:01:34 -0500
re:
http://www.garlic.com/~lynn/2010b.html#4 Happy DEC-10 Day
http://www.garlic.com/~lynn/2010b.html#8 Happy DEC-10 Day

somebody's (naming architecture, osi transition) trip report also mentioning decnet/gosip (from 1989)
http://www.garlic.com/~lynn/2003e.html#71 GOSIP

some mention of wecker & decnet ... also 16% of vm370 burlington mall development group had gone to dec ... rather than moving to pok to support mvs/xa
http://www.garlic.com/~lynn/2006m.html#21 The very first text editor

misc. old email with misc. dec news items (inclucding decnet/osi item)
http://www.garlic.com/~lynn/2006u.html#email880331
in this post
http://www.garlic.com/~lynn/2006u.html#9 The Future of CPUs: What's After Multi-Core?

misc. old email w/various old dec announcements (mostly smp)
http://www.garlic.com/~lynn/2007.html#email880324
and
http://www.garlic.com/~lynn/2007.html#email880329
in this post
http://www.garlic.com/~lynn/2007.html#46 How many 36-bit Unix ports in the old days?

tcp/ip was (internetworking) technology basis for modern internet, nsfnet backbone was original (internetworking) operational implementation, and CIX was business basis for modern internet. misc. past email related to nsfnet backbone
http://www.garlic.com/~lynn/lhwemail.html#nsfnet

we had an internal high-speed backbone and were working with various parties that were working up to nsfnet backbone. We could demonstrated T1 and higher speed operation ... and believe that was part of reason that the nsfnet backbone RFP specified T1. however, for various internal politics, we weren't allowed to bid on nsfnet backbone. The director of NSF tried to help the situation by writing a letter to the corporation, copying the ceo ... but that actually aggravated the internal politics (little statements that what we already had running was at least five years ahead of all bid submissions). The winning nsfnet bid ... actually put in 440kbit links (not T1) ... but they had T1 trunks with telco multiplexors (possibly as effort to meet the letter of the rfp). We made some snide references that they might even claim T5, since possibly some of the T1 trunks possibly were in turn multiplexed by telco over T5 trunks.

the next round was nsfnet t3 backbone rfp. there was internal corporate gathering to answer the rfp ... i was the redteam ... there was 20-30 people from half dozen or so labs from around the world on the blue team. at the final review, i presented first ... then the blue team presentation started. five minutes into the blue team presentation ... the person running the review pounded on the table and said he would lie down in front of garbage truck before he let any but the blue team proposal go forward.

past posts mentioning nsfnet
http://www.garlic.com/~lynn/subnetwork.html#nsfnet

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Happy DEC-10 Day

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Happy DEC-10 Day
Newsgroups: alt.sys.pdp10, alt.folklore.computers
Date: Tue, 12 Jan 2010 23:42:08 -0500
re:
http://www.garlic.com/~lynn/2010b.html#7 Bookshelves under BookManager

as mentioned in this
http://www.garlic.com/~lynn/2010b.html#6 Bookshelves under BookManager

POO in the past was cms "script" ... that was actually subset of the architecture book (at one time distributed in RED 3-ring binder ... and called the "red book" ... different from the current public manuals called "red books"). cms command line options to script command would generate/format the full "red book" ... or just the POO subset. the advantage of having a single document was that it helped keep the information in sync. the non-POO part tended to be as large as the POO subset ... and included things like "engineering notes" (implementation considerations on different machines) and detailed instruction justification description.

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Korean bank Moves back to Mainframes (...no, not back)

From: lynn@GARLIC.COM (Anne & Lynn Wheeler)
Subject: Re: Korean bank Moves back to Mainframes (...no, not back)
Newsgroups: bit.listserv.ibm-main
Date: 13 Jan 2010 07:56:21 -0800
Howard Brazee <howard.brazee@cusys.edu> writes:
And the IS community has to realize that any solution is flawed if it requires these salesmen and/or everybody who does on-line shopping to be experts in security.

we had been called in to consult with a small client/server startup that wanted to do payment transactions on their server ... the startup had also invented this technology called "SSL" they wanted to use. Part of the effort was deploying something called a "payment gateway" (we periodically claim is the original SOA) ... misc. past posts
http://www.garlic.com/~lynn/subnetwork.html#gateway

the effort is now frequently called "electronic commerce". given the ease that crooks can harvest account numbers and use them for fraudulent transactions ... I drew up a list of things required for commerce servers enabled for payment transactions ... like all individuals involved in any way needed to have FBI background checks (type required of individuals in sensitive positions at financial institutions). part of this was that long term numbers claim that insiders are involved in 70% of such events.

related comments about current paradigm in threads about "naked transactions"
http://www.garlic.com/~lynn/subintegrity.html#payments

somewhat as the result of the work on "electronic commerce", in the mid-90s, we were invited to participate in the x9a10 financial standard working group which had been given the requirement to preserve the integrity of the financial infrastructure for ALL retail payments. as part of that activity there was detailed end-to-end threat & vulnerability studies done of different kinds & modes of retail payments.

x9a10 financial standard working group produced a payment standard that slightly tweaked the paradigm and eliminate the threat and vulnerability from having account numbers and/or other transaction information revealed ... for ALL retail payments (point-of-sale, face-to-face, unattended, credit, debit, internet, ACH, stored-value, aka ALL).
http://www.garlic.com/~lynn/x959.html#x959

x9.59 financial standard didn't do anything about hiding or encrypting the information in transactions ... but eliminated the ability of the crooks being able to use that information for fraudulent transactions.

Now the major use of "SSL" in the world today is this earlier "electronic commerce" work to hide account numbers and transaction details. A side effect of x9.59 financial standard eliminates the need for that hiding and therefor the major use of "SSL" in the world today.

past posts in this thread:
http://www.garlic.com/~lynn/2010.html#71 Korean bank Moves back to Mainframes (...no, not back)
http://www.garlic.com/~lynn/2010.html#72 Korean bank Moves back to Mainframes (...no, not back)
http://www.garlic.com/~lynn/2010.html#73 Korean bank Moves back to Mainframes (...no, not back)
http://www.garlic.com/~lynn/2010.html#77 Korean bank Moves back to Mainframes (...no, not back)
http://www.garlic.com/~lynn/2010.html#93 Korean bank Moves back to Mainframes (...no, not back)
http://www.garlic.com/~lynn/2010.html#95 Korean bank Moves back to Mainframes (...no, not back)
http://www.garlic.com/~lynn/2010.html#96 Korean bank Moves back to Mainframes (...no, not back)
http://www.garlic.com/~lynn/2010.html#97 Korean bank Moves back to Mainframes (...no, not back)
http://www.garlic.com/~lynn/2010.html#98 Korean bank Moves back to Mainframes (...no, not back)
http://www.garlic.com/~lynn/2010b.html#1 Korean bank Moves back to Mainframes (...no, not back)
http://www.garlic.com/~lynn/2010b.html#2 Korean bank Moves back to Mainframes (...no, not back)
http://www.garlic.com/~lynn/2010b.html#3 Korean bank Moves back to Mainframes (...no, not back)

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Larrabee delayed: anyone know what's happening?

Refed: **, - **, - **
From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Larrabee delayed: anyone know what's happening?
Newsgroups: comp.arch
Date: Wed, 13 Jan 2010 11:28:38 -0500
"Steven G. Johnson" <stevenj@alum.mit.edu> writes:
What you're referring to is that at one point we found performance improvements (I believe on an IBM RS/6000, if I remember correctly), by inserting padding into the middle of a *one-dimensional* array (again to avoid cache conflicts), but it didn't seem like there was a sane interface for specifying a 1d array with padding in the middle.

we were brought in to look at issues with one of the major airline res. systems .... first looking at "routes" (basically finding flt(s) from origin to destination) ... that accounted for 25% of workload on large collection of mainframes.

i redid the paradigm and implementation on rs/6000 ... as part of trying to get at least ten times performance improvement. Initial pass got twenty times performance improvement ... then some careful tuning of cache line considerations got another factor of five times improvement (100 times improvement overall). I added in bunch of new features (including collapsing several human interactions into single operation) ... which then brought overall performance of that single interaction back down to about ten times (but it was eliminating several additional interactions/transactions).

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

360 programs on a z/10

Refed: **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **
From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: 360 programs on a z/10
Newsgroups: bit.listserv.ibm-main, alt.folklore.computers
Date: Thu, 14 Jan 2010 08:55:27 -0500
Anne & Lynn Wheeler <lynn@garlic.com> writes:
2nd hand tale of some of the competitors testimony in gov./ibm anti-trust trial ... all computer manufacturers knew by the late 50s that the single most important factor in the market place was to have a compatible architecture across the whole machine line ... and they weren't able to get all the different plant managers to toe the line ... different plant managers responsible for different models would do various optimizations for the particular technology that they were using. Only Watson prevailed in forcing all the plant managers (responbile for the different models) to toe the 360 architecture comptability line.

re:
http://www.garlic.com/~lynn/2010.html#45 360 programs on a z/10

so some number of business reporters from the auto show this week commented that the us auto industry is making statements that they need to be agile and adaptable to react to changing consumer preferences and market conditions in order to compete with foreign competitors.

in the 90/91 time-frame the us auto industry had C4 task force meetings about how to become more profitable and compete with foreign competitors and invited some number of technology vendors to participate. they detailed that big inhibitor was the long product cycle (from idea to rolling off the line) of 7-8 yrs ... when the foreign competitors had cut their cycle to 3-4 yrs and looked to be in process of cutting that in half (18-24 months). the industry had a bunch of details on what was needed ... as well as looking to technology vendors for help in improving the process as well as cutting the elapsed time. One of the examples used was corvette design which tended to have very tight space/size tolerances ... and between initial design and actually starting to manufacture ... several components would have changed size and shape ... and no longer fit (requiring expensive redesign & delay).

I chided some of the pok/mainframe attendees about it might be difficult for them to offer advice, since at the time, they were in similar product cycle situation.

In any case, while in the C4 meetings it was possible for them to clearly articulate all the problems and what all the changes that were required (including being more agile and adaptable), they didn't seem to be actually able to do anything ... all the major stakeholders seemed to have vested interest in preserving the status quo.

misc past posts mentioning Boyd &/or OODA-loops (OODA-loops being one of the best paradigms for characterizing agile and adaptable ... especially in competitive situations; in the past, I had sponsored Boyd's briefings at IBM):
http://www.garlic.com/~lynn/subboyd.html

misc. past posts mentioning C4 task force meetings:
http://www.garlic.com/~lynn/2000f.html#41 Reason Japanese cars are assembled in the US (was Re: American bigotry)
http://www.garlic.com/~lynn/2000f.html#43 Reason Japanese cars are assembled in the US (was Re: American bigotry)
http://www.garlic.com/~lynn/2001k.html#36 Newbie TOPS-10 7.03 question
http://www.garlic.com/~lynn/2003i.html#61 TGV in the USA?
http://www.garlic.com/~lynn/2004h.html#22 Vintage computers are better than modern crap !
http://www.garlic.com/~lynn/2006.html#43 Sprint backs out of IBM outsourcing deal
http://www.garlic.com/~lynn/2006c.html#14 Program execution speed
http://www.garlic.com/~lynn/2006g.html#20 The Pankian Metaphor
http://www.garlic.com/~lynn/2006m.html#49 The Pankian Metaphor (redux)
http://www.garlic.com/~lynn/2007g.html#29 The Perfect Computer - 36 bits?
http://www.garlic.com/~lynn/2007g.html#34 U.S. Cedes Top Spot in Global IT Competitiveness
http://www.garlic.com/~lynn/2007i.html#13 U.S. Cedes Top Spot in Global IT Competitiveness
http://www.garlic.com/~lynn/2007j.html#33 IBM Unionization
http://www.garlic.com/~lynn/2007n.html#31 IBM obsoleting mainframe hardware
http://www.garlic.com/~lynn/2007q.html#4 Horrid thought about Politics, President Bush, and Democrats
http://www.garlic.com/~lynn/2008c.html#22 Toyota Beats GM in Global Production
http://www.garlic.com/~lynn/2008c.html#68 Toyota Beats GM in Global Production
http://www.garlic.com/~lynn/2008e.html#30 VMware signs deal to embed software in HP servers
http://www.garlic.com/~lynn/2008e.html#31 IBM announced z10 ..why so fast...any problem on z 9
http://www.garlic.com/~lynn/2008f.html#50 Toyota's Value Innovation: The Art of Tension
http://www.garlic.com/~lynn/2008h.html#65 Is a military model of leadership adequate to any company, as far as it based most on authority and discipline?
http://www.garlic.com/~lynn/2008i.html#31 Mastering the Dynamics of Innovation
http://www.garlic.com/~lynn/2008k.html#2 Republican accomplishments and Hoover
http://www.garlic.com/~lynn/2008k.html#50 update on old (GM) competitiveness thread
http://www.garlic.com/~lynn/2008k.html#58 Mulally motors on at Ford
http://www.garlic.com/~lynn/2008m.html#21 Fraud due to stupid failure to test for negative
http://www.garlic.com/~lynn/2008m.html#52 Are family businesses unfair competition?
http://www.garlic.com/~lynn/2008n.html#4 Michigan industry
http://www.garlic.com/~lynn/2008p.html#77 Tell me why the taxpayer should be saving GM and Chrysler (and Ford) managers & shareholders at this stage of the game?
http://www.garlic.com/~lynn/2008q.html#22 Is Pride going to decimate the auto Industry?
http://www.garlic.com/~lynn/2009f.html#20 What is the real basis for business mess we are facing today?
http://www.garlic.com/~lynn/2009i.html#2 China-US Insights on the Future of the Auto Industry
http://www.garlic.com/~lynn/2009i.html#3 IBM interprets Lean development's Kaizen with new MCIF product
http://www.garlic.com/~lynn/2009i.html#10 64 Cores -- IBM is showing a prototype already
http://www.garlic.com/~lynn/2009i.html#31 Why are z/OS people reluctant to use z/OS UNIX?
http://www.garlic.com/~lynn/2009o.html#67 I would like to understand the professional job market in US. Is it shrinking?

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

security and online banking

Refed: **, - **, - **, - **
From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: security and online banking
Newsgroups: comp.os.linux.security
Date: Thu, 14 Jan 2010 09:16:42 -0500
Steve Hayes <steve@red.honeylink.blue.co.uk> writes:
I don't think they will be using any sort of public key system for this for three reasons:

1 - implementing public-key encryption and decryption needs more processing power than they would want to put on a chip-and-pin card.

2 - public key would only be useful if the system is both encrypting and decrypting. An entire block of encrypted data would have to be input to the handheld device by the user and perhaps copied back. With RSA, this would be well over 100 digits. I believe there are other public key systems that are better but not enough to make this practical.

3 - public key techniques aren't needed for this job.


actually quite a bit of work was done on that in the 90s, as part of the x9a10 financial transaction standard working group (had been given the requirement to preserve the integrity of the financial infrastructure for ALL retail payments).

we had been called in to consult with a small client/server startup that wanted to do payment transactions on their server ... the startup had also invented this technology called SSL they wanted to use. Part of the effort was deploying something called a "payment gateway" (we periodically claim is the original SOA) ... misc. past posts
http://www.garlic.com/~lynn/subnetwork.html#gateway

the effort is now frequently called "electronic commerce". given the ease that crooks can harvest account numbers and use them for fraudulent transactions ... I drew up a list of things required for commerce servers enabled for payment transactions ... like all individuals involved in any way needed to have FBI background checks (type required of individuals in sensitive positions at financial institutions). part of this was that long term numbers claim that insiders are involved in 70% of such events.

related comments about current paradigm in threads about "naked transactions"
http://www.garlic.com/~lynn/subintegrity.html#payments

somewhat as the result of the work on "electronic commerce", in the mid-90s we were invited to participate in the x9a10 financial standard working group. as part of that activity there was detailed end-to-end threat & vulnerability studies done of different kinds & modes of retail payments. the ALL was things like point-of-sale, attended, unattended, credit, debit, stored-value, gift card, contact, contactless, internet, wireless, transit turnstile, aka ALL (as well as most online banking transactions). The transit industry had requirement that the operation be able to be performed in the limited power (of contactless) and elapsed time (small subsecond) of transit turnstile ... and also be very inexpensive.

I had semi-facetiously joked that I would take a $500 milspec part and aggresively cost reduce by 2-3 orders of magnitude while improving on the security (while also being able to satisfy the transit turnstile contactless power and elapsed time requirements ... as well as cost requirements).

In any case, x9a10 slightly tweaked the paradigm for x9.59 standard
http://www.garlic.com/~lynn/x959.html#x959

and also eliminated the requirement for hiding the account number and transaction detail.

Now, the largest use of SSL in the world today is the previous work related to "electronic commerce" for transaction encryption as part of hiding account number and transaction detail ... however, x9.59 standard eliminates the need to hide that information ... and so would also eliminate the major use of SSL in the world today.

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

How long for IBM System/360 architecture and its descendants?

Refed: **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **
From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: How long for IBM System/360 architecture and its descendants?
Newsgroups: alt.folklore.computers
Date: Thu, 14 Jan 2010 10:55:33 -0500
hancock4 writes:
The descendants of IBM's System/360 architecture, now called the "Z" series, will very likely hit their 50th anniversary in 2014. There's still a huge installed base in service that isn't going anywhere, despite the ongoing conversions to 'client-server' processing.

But long will it live on? I don't think anyone would've expected in 1964 that it would've lasted this long. Will it make it to its 75th anniversary?


one of the biggest users are (at least) the F&I from "FIRE" industry ... recent reference to "FIRE"
http://www.garlic.com/~lynn/2010.html#81 Happy DEC-10 Day

in the 90s ... there was big effort (billions spent just by various institutions in manhatten) to rewrite major applications to eliminate the overnight batch window and move to straight-through processing ... leveraging large numbers of "killer micros". There were large disaster ... since the software technology being used introduced factor of 100* greater overhead (compared to the cobol batch), totally swamping any aniticipated thruput improvement.

a couple yrs ago, I was involved in proposal to industry group for a new effort at straight-through processing ... but using technology that was possibly only 2-3 less efficient (than cobol batch) in achieving highly parallel operation for straight-through processing (taking each individual transaction to completion ... rather than deferring major portion of transaction to overnight batch processing, easily achieving all of the original objectives). the response was nearly scalded cat reaction ... because so many organizations had been so badly burned by the failed 90s efforts.

for any change ... appears to at least require a new generation ... as well as demonstratable cost/benefit .... i.e. cost of rewrite/move has to demonstrate ROI benefit compared to current implementation (including confidence in any new implementation being at least as dependable as implementation that has been running reliably for 20-30 yrs).

90s was also period of growth and they needed both new function as well as additional capacity (part of motivation for change). it may be some time before such a growth period returns. growth is going on in other areas of the world ... which have opportunity for doing new implementations from scratch and not having any consideration regarding legacy stuff.

past posts mentioning failed 90s straight-through processing forey
http://www.garlic.com/~lynn/2007e.html#31 Quote from comp.object
http://www.garlic.com/~lynn/2007l.html#15 John W. Backus, 82, Fortran developer, dies
http://www.garlic.com/~lynn/2007l.html#20 John W. Backus, 82, Fortran developer, dies
http://www.garlic.com/~lynn/2007m.html#36 Future of System/360 architecture?
http://www.garlic.com/~lynn/2007u.html#19 Distributed Computing
http://www.garlic.com/~lynn/2007u.html#21 Distributed Computing
http://www.garlic.com/~lynn/2007u.html#37 folklore indeed
http://www.garlic.com/~lynn/2007u.html#44 Distributed Computing
http://www.garlic.com/~lynn/2007u.html#61 folklore indeed
http://www.garlic.com/~lynn/2007v.html#19 Education ranking
http://www.garlic.com/~lynn/2007v.html#27 folklore indeed
http://www.garlic.com/~lynn/2007v.html#64 folklore indeed
http://www.garlic.com/~lynn/2007v.html#69 Controlling COBOL DDs named SYSOUT
http://www.garlic.com/~lynn/2007v.html#72 whats the world going to do when all the baby boomers retire
http://www.garlic.com/~lynn/2007v.html#81 Tap and faucet and spellcheckers
http://www.garlic.com/~lynn/2008b.html#74 Too much change opens up financial fault lines
http://www.garlic.com/~lynn/2008c.html#92 CPU time differences for the same job
http://www.garlic.com/~lynn/2008d.html#30 Toyota Sales for 2007 May Surpass GM
http://www.garlic.com/~lynn/2008d.html#31 Toyota Sales for 2007 May Surpass GM
http://www.garlic.com/~lynn/2008d.html#73 Price of CPU seconds
http://www.garlic.com/~lynn/2008d.html#87 Berkeley researcher describes parallel path
http://www.garlic.com/~lynn/2008d.html#89 Berkeley researcher describes parallel path
http://www.garlic.com/~lynn/2008g.html#55 performance of hardware dynamic scheduling
http://www.garlic.com/~lynn/2008h.html#50 Microsoft versus Digital Equipment Corporation
http://www.garlic.com/~lynn/2008h.html#56 Long running Batch programs keep IMS databases offline
http://www.garlic.com/~lynn/2008p.html#26 What is the biggest IT myth of all time?
http://www.garlic.com/~lynn/2008p.html#30 Automation is still not accepted to streamline the business processes... why organizations are not accepting newer technolgies?
http://www.garlic.com/~lynn/2008r.html#7 If you had a massively parallel computing architecture, what unsolved problem would you set out to solve?
http://www.garlic.com/~lynn/2009.html#87 Cleaning Up Spaghetti Code vs. Getting Rid of It
http://www.garlic.com/~lynn/2009c.html#43 Business process re-engineering
http://www.garlic.com/~lynn/2009d.html#14 Legacy clearing threat to OTC derivatives warns State Street
http://www.garlic.com/~lynn/2009f.html#55 Cobol hits 50 and keeps counting
http://www.garlic.com/~lynn/2009h.html#1 z/Journal Does it Again
http://www.garlic.com/~lynn/2009h.html#2 z/Journal Does it Again
http://www.garlic.com/~lynn/2009i.html#21 Why are z/OS people reluctant to use z/OS UNIX?
http://www.garlic.com/~lynn/2009i.html#23 Why are z/OS people reluctant to use z/OS UNIX? (Are settlements a good argument for overnight batch COBOL ?)
http://www.garlic.com/~lynn/2009i.html#26 Why are z/OS people reluctant to use z/OS UNIX?
http://www.garlic.com/~lynn/2009i.html#30 Why are z/OS people reluctant to use z/OS UNIX?
http://www.garlic.com/~lynn/2009i.html#38 Why are z/OS people reluctant to use z/OS UNIX?
http://www.garlic.com/~lynn/2009i.html#43 Why are z/OS people reluctant to use z/OS UNIX? (Are settlements a good argument for overnight batch COBOL ?)
http://www.garlic.com/~lynn/2009i.html#60 In the USA "financial regulator seeks power to curb excess speculation."
http://www.garlic.com/~lynn/2009l.html#57 IBM halves mainframe Linux engine prices
http://www.garlic.com/~lynn/2009m.html#81 A Faster Way to the Cloud
http://www.garlic.com/~lynn/2009n.html#13 UK issues Turning apology (and about time, too)
http://www.garlic.com/~lynn/2009o.html#81 big iron mainframe vs. x86 servers
http://www.garlic.com/~lynn/2009p.html#57 MasPar compiler and simulator
http://www.garlic.com/~lynn/2009q.html#67 Now is time for banks to replace core system according to Accenture
http://www.garlic.com/~lynn/2009r.html#35 70 Years of ATM Innovation
http://www.garlic.com/~lynn/2009r.html#47 70 Years of ATM Innovation
http://www.garlic.com/~lynn/2010.html#77 Korean bank Moves back to Mainframes (...no, not back)

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

How long for IBM System/360 architecture and its descendants?

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: How long for IBM System/360 architecture and its descendants?
Newsgroups: alt.folklore.computers
Date: Thu, 14 Jan 2010 11:48:27 -0500
Stephen Wolstenholme <steve@tropheus.demon.co.uk> writes:
I've no idea how long 360 architecture will live. That sort of prediction is not possible. I do know that 'client-server' and mainframe architecture are not mutually exclusive.

re:
http://www.garlic.com/~lynn/2010b.html#16 How long for IBM System/360 architecture and its descendants?

there was period that mainframe architecture and client/server was perceived as antithetical ... by the mainframe communication group looking at preserving their terminal emulation install base. some past posts
http://www.garlic.com/~lynn/subnetwork.html#emulation

terminal emulation contributed to early uptake of PC ... but later as PCs & PC software became more sophisticated, terminal emulation started to represent an inhibitor. senior technical person from disk division even had a presentation at annual world-wide internal communication conference that started out saying the head of the communication group was going to be responsible for demise of the disk division.

the issue was that terminal emulation was starting to be such a stranglehold on data into/out-of mainframe datacenter ... that data was starting to leak out of the datacenter at increasing alarming rate to reside at locations outside of the (mainframe) datacenter.

in the time-frame we had came up with 3-tier architecture and was out pitching it to customer executives ... and taking lots of barbs from the communication group (part of preserving the terminal emulation install base)
http://www.garlic.com/~lynn/subnetwork.html#3tier

part of the concepts came from dealing with NCAR ... and using mainframe as file/data server (early NAS/SAN) for supercomputers ... i.e. technologies other than terminal emulation ... allowing mainframes to have efficient & high-thuput connectivity ... some this was related to my high-speed data transport project
http://www.garlic.com/~lynn/subnetwork.html#hsdt

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

security and online banking

Refed: **, - **, - **, - **, - **, - **, - **, - **
From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: security and online banking
Newsgroups: comp.os.linux.security
Date: Thu, 14 Jan 2010 12:04:26 -0500
re:
http://www.garlic.com/~lynn/2010b.html#15 security and online banking

in mid-90s ... there were various presentations by dial-up online banking organizations. the consumer banking organization were talking about moving to the internet (& using SSL) ... a major reason was that it offloaded the enormous customer support costs for (serial-port) dialup modems to the internet service providers (presentations about some operations having libraries having greater than 60 different software drivers for their dialup banking software, and large customer call center support costs). side-issue was that the ISPs could amortise all of that support across a much larger market ... rather than just dialup online banking (and as it turns out, the much larger internet market, prompted vendors to work out lots of the serial-port dial-up modem problems and started including tested support in original product ... instead of it being an aftermarket problem).

however, the cash-management, commercial/business dialup online banking operations said that they would never move to internet (even if they got 128-bit SSL instead of 40-bit SSL) ... because of a long list of threats and vulnerabilities; nearly every possibly kind of exploit that has occured in the past 15 yrs was already on their list in the mid-90s (as reasons for not moving to the internet).

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

STEM crisis

Refed: **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **
From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: STEM crisis
Newsgroups: alt.folklore.computers
Date: Thu, 14 Jan 2010 19:03:59 -0500
US faces critical lack of (mad) computer scientists
http://www.theregister.co.uk/2010/01/13/darpa_mad_scientist_base_boost_plan/

from above:
The downward trend in college graduates with STEM [science, technology, engineering and maths] majors is particularly pronounced in Computer Science (CS).

... snip ...

gov. & large institutions have identified this as a major institutional risk for some time (along with the baby boomers retiring and not enough replacements waiting in the wings).

and then there is this from today ...

Obama convenes summit to modernize government technology
http://www.federaltimes.com/article/20100114/IT03/1140304/1055/AGENCY

when Obama came into office, he made several statements about taking back the gov ... because so much of the gov. had been outsourced to vendors ... who placed their own interests ahead of the gov.

This goes along with past article about Success of Failure ... vendors finding that it is more profitable to have a series of failed projects (especially noticeable are whole string of failed fed. gov. IT/dataprocessing modernization projects) than having a success.

A couple past references
http://www.garlic.com/~lynn/2009o.html#25 Opinions on the 'Unix Haters' Handbook'
http://www.garlic.com/~lynn/2009o.html#41 U.S. house decommissions its last mainframe, saves $730,000
to article Success of Failure
http://www.govexec.com/management/management-matters/2007/04/the-success-of-failure/24107/

a little x-over with recent failed "overnight batch" modernization post
http://www.garlic.com/~lynn/2010b.html#16 How long for IBM System/360 architecture and its descendants?

a few past references to the STEM "crisis":
http://www.garlic.com/~lynn/2006h.html#38 Taxes
http://www.garlic.com/~lynn/2007l.html#5 IBM Unionization
http://www.garlic.com/~lynn/2008.html#57 Computer Science Education: Where Are the Software Engineers of Tomorrow?
http://www.garlic.com/~lynn/2008b.html#6 Science and Engineering Indicators 2008
http://www.garlic.com/~lynn/2008c.html#69 Toyota Beats GM in Global Production
http://www.garlic.com/~lynn/2008j.html#41 NSFnet -- 20 Years of Internet Obscurity and Insight
http://www.garlic.com/~lynn/2008l.html#0 IBM-MAIN longevity

a few past past posts mentioning failed gov. modernization
http://www.garlic.com/~lynn/2007e.html#52 US Air computers delay psgrs
http://www.garlic.com/~lynn/2008m.html#41 IBM--disposition of clock business
http://www.garlic.com/~lynn/2008m.html#45 IBM--disposition of clock business
http://www.garlic.com/~lynn/2009q.html#29 Check out Computer glitch to cause flight delays across U.S.
http://www.garlic.com/~lynn/2009q.html#31 Check out Computer glitch to cause flight delays across U.S.

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Happy DEC-10 Day

Refed: **, - **, - **
From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Happy DEC-10 Day
Newsgroups: alt.sys.pdp10, alt.folklore.computers
Date: Fri, 15 Jan 2010 09:55:37 -0500
jmfbahciv <jmfbahciv@aol> writes:
He didn't balance the budget.

the folklore was that in the mid-90s there was still a gap being able to claim the budget had been "balanced". this was in the period when they were auctioning off airwaves for large sums of money (also before the internet bubble burst). the strategy to close the gap was congress passes legislation to mandate HDTV ... the HDTV digital transmission uses less airwaves than the old analog ... then the free'ed-up airwaves are auctioned off for enormous amounts of money ... sufficient to close the gap for balanced budget.

a couple post posts mentioning HDTV & balanced budget folklore
http://www.garlic.com/~lynn/2008.html#43 dig. TV
http://www.garlic.com/~lynn/2008f.html#61 Primaries (USA)

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

security and online banking

Refed: **, - **, - **, - **, - **, - **, - **, - **, - **, - **
From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: security and online banking
Newsgroups: comp.os.linux.security
Date: Fri, 15 Jan 2010 11:19:54 -0500
Steve Hayes <steve@red.honeylink.blue.co.uk> writes:
The bank sends a challenge of some number of digits which the user types into the handheld device. Enough digits are used to make it unlikely that the same challenge would ever be sent twice.

re:
http://www.garlic.com/~lynn/2010b.html#15 security and online banking
http://www.garlic.com/~lynn/2010b.html#18 security and online banking

old posts (in sci.crypt) driving down to visit the company that makes the devices (halfway between amsterdam and brussels) ... and then driving on down to brussels for EU FINREAD standards meeting
http://www.garlic.com/~lynn/2001g.html#57 Internet banking
http://www.garlic.com/~lynn/2001g.html#60 PKI/Digital signature doesn't work

finread URL reference in the above has gone 404, but wayback machine can be your friend
http://web.archive.org/web/20011123013940/http://www.cenorm.be/isss/workshop/finread/

finread from the later part of the 90s was to address significant portion of the things that the cash management/commercial dialup online banking operations highlighted as major vulnerabilities ... including whole slew of end-point compromises of PC (in some sense ... countermeasure was to move the end-point to the finread device).

finread got caught up in the disaster of some hardware token deployments from the period and the resulting widely spread opinion in the financial industry that chipcards weren't practical in the consumer market.

the deployment disasters turned out not to be with the actual hardware tokens ... but with the card acceptor devices (card readers) that were given way as part of the programs. it appeared that they got a lot of obsolete serial-port devices (for the give-away) and ran into the enormous customer support issues/problems that had earlier motivated dial-up online banking to move to the internet (apparently the ephemeral financial infrastructure institutional knowledge about serial-port support problems had evaporated in the few short years between move of consumer online banking from proprietary dial-up to the internet ... and the smartcard deployment programs).

misc. past posts mentioning finread
http://www.garlic.com/~lynn/subintegrity.html#finread

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

How long for IBM System/360 architecture and its descendants?

Refed: **, - **, - **
From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: How long for IBM System/360 architecture and its descendants?
Newsgroups: alt.folklore.computers
Date: Fri, 15 Jan 2010 18:24:25 -0500
Peter Flass <Peter_Flass@Yahoo.com> writes:
I argue that there's no reason for a RISC architecture to be full of "gotchas" and exceptions to trip up the programmer, that's just sloppy work by the designers.

I've periodically claimed that John did RISC to go to the opposite extreme from the extreme complexity of future system effort ... misc. past posts
http://www.garlic.com/~lynn/submain.html#futuresys

future system was going to completely replace 360 ... and was as (at least as) different from 360 as 360 had been from prior generations.

in mid-70s technology meeting with risc presentation ... there were lots of references that whenever there was trade-off between hardware complexity and software complexity ... decision was hardware simplicity ... and software complexity would be used to enable simpler hardware.

for instance there was no cache consistency between I-cache (instruction) and (store-in) D-cache (data). This resulted in scenarios where a loader would be operating on instruction image brought into memory ... and any modifications would appear in D-cache ... and wouldn't necessarily be in main memory for fetch by I-cache.

To make it work ... loader had to execute instructions that forced modified data from D-cache to memory and if the corresponding memory locations were in the I-cache, they were invalidated.

A whole lot of technology went into the 801/risc pl8 programming language ... which was designed to compensate for 801/risc hardware shortcomings/simplicity.

I don't anybody ever anticipated that C would ever be used for programming risc processor.

misc past posts mentioning 801, risc, iliad, romp, rios, power, power/pc, etc
http://www.garlic.com/~lynn/subtopic.html#801

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Happy DEC-10 Day

Refed: **, - **, - **
From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Happy DEC-10 Day
Newsgroups: alt.sys.pdp10, alt.folklore.computers
Date: Sat, 16 Jan 2010 00:44:37 -0500
ArarghMail001NOSPAM writes:
IIRC, I used that gateway to get some of the RFCs. That would have been in the very late 80s or very early 90s. Cost extra too, IIRC.

i started keeping shadow of RFCs possibly 85 or 86 or so (ftp.nisc.sri.com) ... then started do index of rfcs ... current
http://www.garlic.com/~lynn/rfcietff.htm

... and consistancy checking of RFC process. some of it started showing up as section 6.10 in STD1 (starting about 1600 or so) for a time.

i expanded to doing filelist of several dozen sites on regular basis and getting copies of all sorts of stuff from late 80s and early 90s ... some of which I still have ... for example following files from policies directory:


MRNet Brief.txt
README
Template-albany.edu
Template-arizona.edu
Template-auburn.edu
Template-auburnschl.edu
Template-bates.edu
Template-bowdoin.edu
Template-brown.edu
Template-bu.edu
Template-bvc.edu
Template-carleton.edu
Template-cba.uga.edu
Template-ceu.edu
Template-cookman.edu
Template-ctstateu.edu
Template-dartmouth.edu
Template-eiu.edu
Template-exploratorium.edu
Template-hamptonu.edu
Template-iastate.edu
Template-macalstr.edu
Template-mccc.edu
Template-miu.edu
Template-mnsmc.edu
Template-muskingum.edu
Template-mwc.edu
Template-ncsu.edu
Template-nevadanet
Template-niu.edu
Template-noao.edu
Template-provo.edu
Template-ricks.edu
Template-rpi.edu
Template-sl.edu
Template-snc.edu
Template-spc.edu
Template-spu.edu
Template-sru.bitnet
Template-stmarys-ca.edu
Template-suffolk.edu
Template-susqu.edu
Template-tarleton.edu
Template-trinity.edu
Template-twu.edu
Template-ua.edu
Template-uidaho.edu
Template-uiowa.edu
Template-umass.edu
Template-unf.edu
Template-uoregon.edu
Template-uwgb.edu
Template-vims.edu
Template-westga.edu
Template-wlu.edu
Template-wofford.edu
Template-wooster.edu
Template.umsl.edu
Template.uncecs.edu
[mail]mail.gateway
[mail]ua-standards
[uofa.commandments]commandments.version-2[obsolete]
[uofa]assigned-numbers-and-names
ans.policy
barrnet.policy
bowdoin-computer-use-policy
cerfnet.policy
cicnet.policy
cren.policy
dartmouth-computing-code
eiu-policy
farnet.policy
fricc.policy
indnet.policy
jvnc.policy
los-nettos.policy
ls-lsR
michnet.policy
mnsmc-policy
mrnet.0README
mrnet.policy
ncsanet.policy
nearnet.policy
netpolicy.src
nevadanet.policy
northwestnet.policy
nsfnet.policy
nysernet.policy
oarnet.policy
onet.policy
prepnet.policy
rpi-rcs-conditions-of-use
spu-internet-user-guide
statement-of-computer-ethics
uiowa-internet-man-page
uninet.policy
uonet-access
uonet-appendixes-glossary
uonet-dos-ps-network-workstation
uonet-guidelines
uonet-mac-network-workstation
uonet-networked-unix-workstations
uonet-users-guide
uonet-vax-vmx-network-software
usenet-news-policy
widener-student-computing-resource-policy

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Happy DEC-10 Day

Refed: **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **
From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Happy DEC-10 Day
Newsgroups: alt.sys.pdp10, alt.folklore.computers
Date: Sat, 16 Jan 2010 11:24:47 -0500
William Hamblen <william.hamblen@earthlink.net> writes:
At the moment OASDI taxes are roughly 15% of wages, and pensions paid are roughly equal to taxes collected. If nothing changes on the pension side, the tax side will have to be increased to 30% of wages quite soon. The Congress could break the connection between wages earned and pension paid and make the pension means tested in order to avoid big tax inceases. What that will do for retirement saving I can't predict. A lot of people will figure why save anything if the government will just reduce benefits.

there was article that baby boomer generation is four times larger than the previous generation and that the generation following the baby boomers is only half as large ... as the baby boomer bubble moves into retirement ... the ratio between the following working generation to retiree (baby boomer) generation is reduced by factor of eight times (1/8th ratio in the following generation in prime earning to the number of retirees).

during the baby boomer prime earning years ... the SS collections was larger than the SS payouts ... the excess going into general fund (folklore is bottom desk drawer somewhere in west virginia with the IOUs) ... effectively turning SS into "pay as you go" retirement fund (as opposed to fully funded retirement) ... and used for underwriting general federal expenditures.

with a change of factor eight in the ratio of those paying in to those receiving benefits ... there is some implication that the 15percent will have to increase 120percent ... to maintain the same level of benefits (because the increase in the ratio of those receiving benefits to the number of working and paying taxes).

there had been some past legislation from congress requiring corporations to move to fully funded retirement plans (i.e. money paid in as person worked was what was used to pay their benefits) ... some number of corporations that were on pay-as-you-go ... found that it required huge increase (they were also leveraging the baby boomer worker generation was so much larger than the retired generation). some number just declared bankruptcy and threw their workers into the gov. pension plan.
http://www.pbgc.gov/
and
http://www.pionline.com/article/20090406/PRINTSUB/304069981

however, in the mid-90s, the fully funded retirement plans were put at risk ... when legislation allowed the pension plans to be listed as a corporate asset (potentially making them vulnerable to creditors). some discussion of issues:
http://www.nysscpa.org/cpajournal/2008/308/essentials/p32.htm

there were some articles that CEOs made their stock value bonus objectives solely based on that change to corporate assets.

the other issue is that there is lots of reports that the following generation is less well educated and less qualified for well-paying-jobs ... so the base income being taxed is drastically reduced ... in order to maintain same level of benefits ... it may be necessary to increase SS tax to 250% (or more).

recent reference to ongoing declining education level:
http://www.garlic.com/~lynn/2010b.html#19 STEM crisis

misc. past posts mentioning the worker/retiree generation ratio change:
http://www.garlic.com/~lynn/2008i.html#98 dollar coins
http://www.garlic.com/~lynn/2008o.html#58 Everyone is getting same deal out of life: babyboomers can't retire but they get SS benefits intact
http://www.garlic.com/~lynn/2009m.html#61 August 7, 1944: today is the 65th Anniversary of the Birth of the Computer
http://www.garlic.com/~lynn/2009o.html#37 Young Developers Get Old Mainframers' Jobs

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Happy DEC-10 Day

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Happy DEC-10 Day
Newsgroups: alt.sys.pdp10, alt.folklore.computers
Date: Sat, 16 Jan 2010 11:38:50 -0500
Morten Reistad <first@last.name> writes:
There was a long transition centered around jan 1st 1987; and that was the intended cut date. It slipped, but not by much.

The Battle of the Commercial Internet happened january 1991 to january 1992, otherwise known as the "cix wars". It was about universal connectivity, and not having cencorship at the core of the Internet, in the form of AUPs, traffic priority etc.


the other scenario for the AUPs was the telcos trying to solve the chicken&egg scenario. fiber &/other new technology drastically increased available bandwidth ... but telcos were finding impossible to solve problem ... to significantly increase the bandwidth use ... the price per bit had to be drastically reduced (even to encourage the developement new generation of bandwidth hungry applications). The telcos have significant fixed run rate ... drastically reducing bandwidth useage rates would have the telcos operating at large loss for several years before new generation of bandwidth hungry applications appeared.

the nsfnet backbone T1 RFP was $11.2M ... however, estimates are that the amount of resources that commercial entities put into the backbone was at least four times that amount. the scenario is that the telcos could have a restricted use free bandwidth sandbox as incubator for next generation of bandwidth hungry applications ... w/o affecting their commercial revenue streams.

that is somewhat separate from some of the technology choices put into that T1 RFP (i.e. 440kbit links ... not real T1 & higher-speed links that we were running) ... recent post
http://www.garlic.com/~lynn/2010b.html#10 Happy DEC-10 Day

reference to list of some of the old AUPs that I still have
http://www.garlic.com/~lynn/2010b.html#23 Happy DEC-10 Day

some old email related to nsfnet backbone
http://www.garlic.com/~lynn/lhwemail.html#nsfnet

misc. past posts related to nsfnet
http://www.garlic.com/~lynn/subnetwork.html#nsfnet

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Happy DEC-10 Day

Refed: **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **
From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Happy DEC-10 Day
Newsgroups: alt.sys.pdp10, alt.folklore.computers
Date: Sat, 16 Jan 2010 12:04:48 -0500
Morten Reistad <first@last.name> writes:
The 2000 boom puzzled us as financial analysts. We applied all the bubble models we had, and they all told us that this should have crashed already. But it kept going, going, going. We still don't know what kept it fuelled for so long. There must have been a lot (a _LOT_) of external liquidity added that excaped the models. I still wonder what happened.

there appeared to be a community of investment bankers that ran IPO mills, put in investment, two years of hype ... and then IPO ... maybe $20m in on the front in, $2b out on the backside. there was even benefit if the new company never actually succeeded ... since that left the market still open for the next IPO.

lots of people buying the stock never actually understood anything about the technology of the company they were buying into ... they just got caught up in the hype. being caught up in the hype of something new & not understood ... resulted in lots of people simply ignoring fundamentals (some flavor of "emperor's new clothes")

new american culture ... Success of Failure ... recent reference:
http://www.garlic.com/~lynn/2010b.html#19 STEM crisis

there has been some claims about similarities with the internet boom and CDOs (as well as S&L crisis).

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Happy DEC-10 Day

Refed: **, - **, - **, - **, - **, - **, - **, - **, - **
From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Happy DEC-10 Day
Newsgroups: alt.sys.pdp10, alt.folklore.computers
Date: Sat, 16 Jan 2010 12:43:21 -0500
Pat Farrell <pfarrell@pfarrell.com> writes:
We ran into that much earlier. There were a number of "block mode" terminals that did forms processing locally, and dumped in all of the characters in a single stream, at whatever the modem had. Using them greatly improved the user experience for boring data entry usages. DEC had a very hard time understanding why anyone would want it, let alone expect it to work.

topaz/3101 "mod2s" ... w/o the ROM upgrade they were effectively very similar to other glass-teletypes, "mod2s" added "block mode" ... aka somewhat more like 3270s (had some early mod1s and got rom image from plant site to burn new mod2 roms).

for home office, i upgraded cdi miniterm at 300 baud ... to 3101 at 1200 baud.

misc old topaz/3101 posts
http://www.garlic.com/~lynn/99.html#69 System/1 ?
http://www.garlic.com/~lynn/2000g.html#17 IBM's mess (was: Re: What the hell is an MSX?)
http://www.garlic.com/~lynn/2001b.html#12 Now early Arpanet security
http://www.garlic.com/~lynn/2001b.html#13 Now early Arpanet security
http://www.garlic.com/~lynn/2001h.html#32 Wanted: pictures of green-screen text
http://www.garlic.com/~lynn/2001m.html#1 ASR33/35 Controls
http://www.garlic.com/~lynn/2001m.html#54 Author seeks help - net in 1981
http://www.garlic.com/~lynn/2003c.html#34 diffence between itanium and alpha
http://www.garlic.com/~lynn/2003c.html#35 diffence between itanium and alpha
http://www.garlic.com/~lynn/2003n.html#7 3270 terminal keyboard??
http://www.garlic.com/~lynn/2004e.html#8 were dumb terminals actually so dumb???
http://www.garlic.com/~lynn/2005p.html#28 Canon Cat for Sale
http://www.garlic.com/~lynn/2005r.html#12 Intel strikes back with a parallel x86 design
http://www.garlic.com/~lynn/2006n.html#56 AT&T Labs vs. Google Labs - R&D History
http://www.garlic.com/~lynn/2006y.html#0 Why so little parallelism?
http://www.garlic.com/~lynn/2006y.html#4 Why so little parallelism?
http://www.garlic.com/~lynn/2006y.html#24 "The Elements of Programming Style"
http://www.garlic.com/~lynn/2006y.html#31 "The Elements of Programming Style"
http://www.garlic.com/~lynn/2007e.html#15 The Genealogy of the IBM PC
http://www.garlic.com/~lynn/2007h.html#39 sizeof() was: The Perfect Computer - 36 bits?
http://www.garlic.com/~lynn/2007k.html#40 DEC and news groups
http://www.garlic.com/~lynn/2007s.html#48 ongoing rush to the new, 40+ yr old virtual machine technology
http://www.garlic.com/~lynn/2007t.html#74 What do YOU call the # sign?
http://www.garlic.com/~lynn/2008l.html#79 Book: "Everyone Else Must Fail" --Larry Ellison and Oracle ???
http://www.garlic.com/~lynn/2008m.html#37 Baudot code direct to computers?
http://www.garlic.com/~lynn/2008m.html#38 Baudot code direct to computers?
http://www.garlic.com/~lynn/2008m.html#51 Baudot code direct to computers?
http://www.garlic.com/~lynn/2008m.html#88 Sustainable Web
http://www.garlic.com/~lynn/2008n.html#51 Baudot code direct to computers?
http://www.garlic.com/~lynn/2008r.html#5 What if the computers went back to the '70s too?
http://www.garlic.com/~lynn/2008s.html#22 IBM PC competitors
http://www.garlic.com/~lynn/2009c.html#30 I need magic incantation for a power conditioner
http://www.garlic.com/~lynn/2009d.html#13 Typewrite repair?
http://www.garlic.com/~lynn/2009g.html#45 Netbooks: A terminal by any other name
http://www.garlic.com/~lynn/2009h.html#32 My Vintage Dream PC
http://www.garlic.com/~lynn/2009j.html#40 My "Green Screen" IBMLink is still working
http://www.garlic.com/~lynn/2009j.html#66 A Complete History Of Mainframe Computing
http://www.garlic.com/~lynn/2009l.html#27 August 7, 1944: today is the 65th Anniversary of the Birth of the Computer
http://www.garlic.com/~lynn/2009n.html#9 Existence of early 360 software ( was Re: Continous Systems Modelling Packa
http://www.garlic.com/~lynn/2009r.html#63 tty
http://www.garlic.com/~lynn/2009s.html#0 tty

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Happy DEC-10 Day

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Happy DEC-10 Day
Newsgroups: alt.sys.pdp10, alt.folklore.computers
Date: Sat, 16 Jan 2010 13:02:40 -0500
Eric Chomko <pne.chomko@comcast.net> writes:
A company as big as DEC doesn't die by one thing. IBM had similar problems as DEC but recovered. DEC could have recovered but made bad decisions to get into trouble and made bad decisions trying to get out of trouble. Give IBM credit for making good decisions to get out of trouble.

also IBM had much broader market. A lot of DEC was in the vax/vms mid-range market ... which endicott 43xxs also sold into. there was big explosion in the mid-range market starting in the late 70s and going up to the mid-80s. 43xx was very sales to vax ... except 43xx also had some huge corporate sales with machines being ordered in several hundred at a time (sort of precursor to departmental servers). endicott expected later 43xx models in the mid-80s to see similar sale volumes that the earlier models saw ... but by that time the mid-range market was moving to workstations and large PCs.

this can be seen in the vax volumes ... old post with decade of vax sales volumes, sliced & diced by year, model, us/non-us:
http://www.garlic.com/~lynn/2002f.html#0 Computers in Science Fiction

it had traumatic effect on endicott sales ... but other parts of the corporation continued to be somewhat unaffected for a time.

endicott machines also accounted for huge explosion in the number of nodes on the internal network ... which was larger than the arpanet/internet from just about the beginning until possibly late 85 or early 86
http://www.garlic.com/~lynn/subnetwork.html#internalnet

in jan '83 when arpanet was making transition to internet (having approx. 100 nodes or 255 hosts ... since arpanet counted IMP network nodes as separate from the number of attached host processors) .. the internal network was getting close to passing 1000 nodes/hosts ... lots of the rapid growth being 43xx machines ... misc. past email mentioning 43xx
http://www.garlic.com/~lynn/lhwemail.html#43xx

one of the issues about internet passing internal network number of nodes/hosts in the mid-80s ... was the appearance of workstations and PCs as nodes ... while the internal network was forced to use terminal emulation for such machines.
http://www.garlic.com/~lynn/subnetwork.html#emulation

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Happy DEC-10 Day

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Happy DEC-10 Day
Newsgroups: alt.sys.pdp10, alt.folklore.computers
Date: Sat, 16 Jan 2010 12:48:18 -0500
jmfbahciv <jmfbahciv@aol> writes:
It was worse than what you describe. TW was expected to write the software of new disks without having one on the system (because product management had promised the first ones to customers; product management had to be retrained every damned time that, in order to deliver the hardware, the software had to exist and that the software couldn't exist until we had the hardware installed on our system for the months required to write, test, and load test the software.

i was being allowed to play disk engineer in bldgs. 14&15 ... because I would help them out with stuff ... and they would let me play with some of their stuff. as part of validating disk operations ... disk labs ... would get early engineering processors ... like maybe 3rd or 4th ... as soon as the processor engineers had extra.

at one point i was doing some work on (disk) engineering 4341 for endicott 4341 software & performance groups ... since I had better access to 4341 (than people in various endicott 4341 groups).

misc. past posts about getting to play disk engineer
http://www.garlic.com/~lynn/subtopic.html#disk

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Happy DEC-10 Day

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Happy DEC-10 Day
Newsgroups: alt.sys.pdp10, alt.folklore.computers
Date: Sat, 16 Jan 2010 14:15:40 -0500
Anne & Lynn Wheeler <lynn@garlic.com> writes:
in jan '83 when arpanet was making transition to internet (having approx. 100 nodes or 255 hosts ... since arpanet counted IMP network nodes as separate from the number of attached host processors) .. the internal network was getting close to passing 1000 nodes/hosts ... lots of the rapid growth being 43xx machines ... misc. past email mentioning 43xx
http://www.garlic.com/~lynn/lhwemail.html#43xx


re:
http://www.garlic.com/~lynn/2010b.html#28 Happey DEC-10 Day

the major internal network technology started out on cp67/cms and moved to vm370/cms ... it was relatively nicely layered ... even with a kind of gateway functionality in every node.

the major mainframe batch system networking appeared to have been inherited from HASP (where the networking source code changes had the identifier "TUCC" in cols. 68-71). HASP/JES had one byte index (255 entry) table to define things. It started out being used for "psuedo" unit record devices (printers, punches readers) ... and typical system might have 60-80 such entries defined. The networking code then utilized the remaining entries to define network nodes i.e. limiting things to under 200 definitions. So effectively for nearly the whole lifetime of the internal network ... MVS/JES nodes were typically limited to edge node (not only would they discard traffic for destination nodes that weren't in their table ... but they would also discard traffic when the originating node wasn't their table).

the other problem was that they had jumbled header information ... mixing up JES control information with network control information. In fact, traffic between JES systems at different release levels was notorious for crashing JES & bringing down production MVS systems.

the main internal networking implementation (on cp67/cms & then vm370/cms) had to develop a set of HASP & then JES "drivers" ... that emulated JES headers to immediately connected JES sysetms. Eventually there was a whole library of JES drivers that not only emulated JES header information ... but if the traffic originated from some other JES system ... it might be necessary to have the psuedo driver reformat various JES fields as countermeasure to MVS system crashes. There was an infamous case of san jose JES/MVS system causing Hursley MVS systems to crash ... and it getting blamed on the vm370 networking code (because the vm370 network code was suppose to have all the reformating implementation from keeping different JES systems causing each other to crash). misc. past posts mentioning hasp, jes, etc
http://www.garlic.com/~lynn/submain.html#hasp

by the time of bitnet (& earn in europe, educational network using technology similar to internal network)
http://www.garlic.com/~lynn/subnetwork.html#bitnet

for customers, were trying to only ship the JES family of drivers with vm370/cms (eliminating shipping the native vm370/cms drivers ... even tho they had higher function and higher thruput ... than the JES stuff). bitnet then started to exceed 200 nodes ... and JES had to finally get around to shipping support for 999 nodes ... but that was well after the internal network had passed 1000 nodes (then JES had to increase the limit to 1999 nodes ... but after when the internal network had already passed 2000 nodes).

numbers from some source circa 1985 (aka "VNET" is the internal corporate network):
BITNET 435 ARPAnet 1155 CSnet 104 (excluding ARPAnet overlap) VNET 1650 EasyNet 4200 UUCP 6000 USENET 1150 (excluding UUCP nodes)

misc. old posts referencing the above counts:
http://www.garlic.com/~lynn/2002k.html#26 DEC eNet: was Vnet : Unbelievable
http://www.garlic.com/~lynn/2006t.html#50 The Future of CPUs: What's After Multi-Core?
http://www.garlic.com/~lynn/2008l.html#2 IBM-MAIN longevity
http://www.garlic.com/~lynn/2008l.html#6 IBM-MAIN longevity
http://www.garlic.com/~lynn/2008l.html#12 IBM-MAIN longevity
http://www.garlic.com/~lynn/2008m.html#18 IBM-MAIN longevity
http://www.garlic.com/~lynn/2008q.html#45 Usenet - Dead? Why?
http://www.garlic.com/~lynn/2008r.html#9 Comprehensive security?
http://www.garlic.com/~lynn/2009j.html#19 Another one bites the dust

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Happy DEC-10 Day

Refed: **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **
From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Happy DEC-10 Day
Newsgroups: alt.sys.pdp10, alt.folklore.computers
Date: Sat, 16 Jan 2010 14:43:32 -0500
Pat Farrell <pfarrell@pfarrell.com> writes:
Yes, the 3270 style was a lot easier on the main CPU, and provided a decent user experience when filling out forms. A lot of vendors made them, including DEC eventually with the VT132/VT131

Using them was very uncomfortable to anyone used to a Tops-10/Tops-20 system, where the computer responded to every character as it was entered.


they were uncomfortable to anybody use to interactive computing.

3277 had a lot of the technology in the computer head ... so it was possible to make some modifications. a particularly annoying feature was if you happen to hit a key and the same time the screen was being written, the keyboard would lock and you have to hit reset to clear it.

one "fix" as a small "FIFO" box ... unplug the keyboard from the head, plug in the FIFO box and then plug the keyboard into the box ... it could queue some number of keystrokes to mask the half-duplex operation. it was also possible to do some soldering inside the keyboard to decrease the repeat key "delay" and the repeat key "rate".

that all changed in the moved to 3274/3278 ... a lot of the electronics were moved back into the 3274 controller ... making the 3278 much cheaper to manufacture ... but further killing it for interactive work.

old 3272/3277 comparison with 3274/3278:


               hardware     TSO 1sec.    CMS .25sec.     CMS .11sec.
3272/3277        .086        1.086         .336            .196
3274/3278        .530        1.530         .78             .64

from this post with much longer discussion:
http://www.garlic.com/~lynn/2001m.html#19 3270 protocol

during late 70s ... somebody from another internal installation was claiming the best timesharing service in the company ... using .25sec system response as example. I pointed out that I was running on similar hardware with nearly identical workload and getting .11sec system response; they then tried to make some statements that it was never fair to compare anything to stuff I did.

later with PCs and terminal emulation ... those with ANR 3277 emulation got significantly higher file upload/download than those stuck with 3278 emulation.

I had done a lot of performance & algorithm stuff as undergraduate that got shipped in cp67 ... but was later dropped in some of the simplification that went on in the morph to vm370. The univ. & SHARE user groups kept lobbying IBM to incorporate my stuff back into vm370 (customer calls for things like the "wheeler" scheduler). I kept doing cp67 & vm370 stuff all during the future system period ... misc. past posts mentioning future system
http://www.garlic.com/~lynn/submain.html#futuresys

since future system was going to completely replace 360/370 ... lots of work stopped on 370. when future system effort was killed, there was mad rush to get stuff back into the 370 product pipeline. that was in large part behind picking up lots of stuff I had been doing and shipping it.

some old email about cp67 & vm370 work in the period:
http://www.garlic.com/~lynn/2006v.html#email731212 731212
http://www.garlic.com/~lynn/2006w.html#email750102 750102
http://www.garlic.com/~lynn/2006w.html#email750430 750430

however, after that short period (where only relatively small amount made it out), it was pretty much return to business as usual and little additional work made it out of internal operation.

one of my hobbies had been building, distributing, and supporting highly modified/enhanced systems for internal use (independent of others picking up changes I made for product ship). at one point, I claimed that I had a (personal) distribution list that was as large as the total number of MULTICS systems that ever shipped.

Some number of the CTSS people had gone to 5th flr of 545tech sq to work on Multics. Others had gone to the science center on 4th flr and did virtual machine systems, internal network technology, interactive computing, inventing GML (precursor to SGML, HTML, XML, etc) ... some past posts mentioning 545tech sq
http://www.garlic.com/~lynn/subtopic.html#545tech

there was a little rivalry between 4th and 5th flrs ... but it wasn't fair to compare the number of MULTICS systems to the total number of customer mainframe (mostly batch) systems, or even much smaller number of customer virtual machine mainframe systems, or the much, much smaller number of internal virtual machine mainframe systems ... however, the number of internal systems I was personally doing was about the same as the total number of MULTICS systems.

misc. past posts mentioning online virtual machine timesharing; internal customers, as well as commercial online timesharing service bureaus
http://www.garlic.com/~lynn/submain.html#timeshare

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Happy DEC-10 Day

Refed: **, - **, - **, - **, - **, - **, - **, - **, - **
From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Happy DEC-10 Day
Newsgroups: alt.sys.pdp10, alt.folklore.computers
Date: Sat, 16 Jan 2010 19:32:14 -0500
Pat Farrell <pfarrell@pfarrell.com> writes:
"response time" was a great source of discussion and more, often turning into marketing speak and into spin and puff.

The key question, usually left unspoken, was "what are you responding to?"

On a TOPS-10 or -20 system one level is responding by echoing the characters as they are typed. Another is responding to COMMND processing of escape, ? or tab. Further up scale are actually doing something, such as responding to a "dir" command.

It gets much more difficult to describe, let alone measure, if you do something "significant" such as issuing a "find fname = 'Pat'" query to your DBMS system. On a well running Tops/10 or 20 system, the 0.25 second response would be typical more business applications.

Obviously, doing a complex sequential search through the tables ( a full table scan in modern DBMS speak) is nearly always going to take much longer.

Meeting a "five second response time, 95% of the time" was a reasonable contract term in the early 80s for large DECsystem-20s.


re:
http://www.garlic.com/~lynn/2010b.html#31 Happy DEC-10 Day

i actually simplified a little. the basis was "trivial" (command) response ... say in the editor and doing some function like finding some string ... frequently involved something on the order of a dozen or so page faults. I also simplified that the organization quoting .25sec was avg. trivial command response ... and I was measuring 95% precentile .11sec response for the same set of commands/operations.

in the wake of demise of future system effort ... i got to release a bunch of stuff for vm370 (lots of it from nearly decade earlier as undergraduate on cp67) ... some amount of it was packaged as separate, special kernel "resource manager" product ... including the "wheeler" fair share scheduler. misc. post mentioning wheeler fair share scheduler (&/or resource manager)
http://www.garlic.com/~lynn/subtopic.html#fairshare

the ".25 sec" response numbers were with my base "resource manager" product ... but not with a lot of other stuff I did subsequently (to get the 95percentile ".11 sec" response).

database was somewhat different ... the original relational/sql implementatation was done on vm/cms 370/145 system ... misc. past posts doing some portion of system/r
http://www.garlic.com/~lynn/submain.html#systemr

and handling some amount of the technology transfer to endicott for what became sql/ds.

later when I was doing cluster scaleup ... we worked with some number of DBMS vendors that had unix implementations as well as vax/cluster implementations. they had a list of things that they felt was done wrong in vax/cluster that I needed to correct. for cluster/scaleup work ... did a cluster distributed lock manager (that addressed the short comings that they believed were in vax/cluster implementation) but provided an API that mimicked vax/cluster (simplifying the port of their vax/cluster implementation to cluster scaleup platform).

old post referencing a jan92 meeting in ellison's conference room
http://www.garlic.com/~lynn/95.html#13

one of the issues that some of the vendors were doing in single complex mode (uniprocessor or smp) was some form of fast commit ... i.e. the transaction considered complete as soon as the log record was written ... but not necessarily the actual buffer record written to DBMS location (recovery after failure then required rerunning log transactions to correctly update dbms records). However, in cluster environment, they would first write record to disk ... if processing was required on a different system. I worked out the details of being able to transfer buffer record, piggybacked on same transmission that transferred lock ownership (effectively doing cache-to-cache copy and preserving fast commit semantics across clustered machines).

the issue for this wasn't so much the actual transfers ... it was working out how to correctly merge the sequence of records from multiple different logs for recovery after a failure.

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Happy DEC-10 Day

Refed: **, - **, - **
From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Happy DEC-10 Day
Newsgroups: alt.sys.pdp10, alt.folklore.computers
Date: Sat, 16 Jan 2010 23:09:15 -0500
Anne & Lynn Wheeler <lynn@garlic.com> writes:
the other scenario for the AUPs was the telcos trying to solve the chicken&egg scenario. fiber &/other new technology drastically increased available bandwidth ... but telcos were finding impossible to solve problem ... to significantly increase the bandwidth use ... the price per bit had to be drastically reduced (even to encourage the developement new generation of bandwidth hungry applications). The telcos have significant fixed run rate ... drastically reducing bandwidth useage rates would have the telcos operating at large loss for several years before new generation of bandwidth hungry applications appeared.

re:
http://www.garlic.com/~lynn/2010b.html#25 Happy DEC-10 Day

following is nsfnet.policy ... from this list of files:
http://www.garlic.com/~lynn/2010b.html#23 Happy DEC-10 Day

[NSFNET] NETUSE.TXT

Interim 3 July 1990
NSFNET
Acceptable Use Policy

The purpose of NSFNET is to support research and education in and among academic institutions in the U.S. by providing access to unique resources and the opportunity for collaborative work.

This statement represents a guide to the acceptable use of the NSFNET backbone. It is only intended to address the issue of use of the backbone. It is expected that the various middle level networks will formulate their own use policies for traffic that will not traverse the backbone.

(1) All use must be consistent with the purposes of NSFNET.

(2) The intent of the use policy is to make clear certain cases which are consistent with the purposes of NSFNET, not to exhaustively enumerate all such possible uses.

(3) The NSF NSFNET Project Office may at any time make determinations that particular uses are or are not consistent with the purposes of NSFNET. Such determinations will be reported to the NSFNET Policy Advisory Committee and to the user community.

(4) If a use is consistent with the purposes of NSFNET, then activities in direct support of that use will be considered consistent with the purposes of NSFNET. For example, administrative communications for the support infrastructure needed for research and instruction are acceptable.

(5) Use in support of research or instruction at not-for-profit institutions of research or instruction in the United States is acceptable.

(6) Use for a project which is part of or supports a research or instruction activity for a not-for-profit institution of research or instruction in the United States is acceptable, even if any or all parties to the use are located or employed elsewhere. For example, communications directly between industrial affiliates engaged in support of a project for such an institution is acceptable.

(7) Use for commercial activities by for-profit institutions is generally not acceptable unless it can be justified under (4) above. These should be reviewed on a case-by-case basis by the NSF Project Office.

(8) Use for research or instruction at for-profit institutions may or may not be consistent with the purposes of NSFNET, and will be reviewed by the NSF Project Office on a case-by-case basis.


... snip ...

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Happy DEC-10 Day

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Happy DEC-10 Day
Newsgroups: alt.sys.pdp10, alt.folklore.computers
Date: Sat, 16 Jan 2010 23:46:40 -0500
re:
http://www.garlic.com/~lynn/2010b.html#33 Happy DEC-10 Day

another random file from the archive:


 [NSFNET] HISTORY.NETCOUNT

This file is a listing by month of the number
of total networks, foreign networks and T3 networks
configured for announcement on the NSFNET
infrastructures during the term of the project.

Month     Total nets      Foreign nets    T3 nets
Jul-88          173             9
Aug-88          217             9
Sep-88          244             9
Oct-88          291            14
Nov-88          313            33
Dec-88          334            33
Jan-89          346            34
Feb-89          384            35
Mar-89          410            38
Apr-89          467            61
May-89          516            95
Jun-89          564            95
Jul-89          603            99
Aug-89          650           137
Sep-89          745           153
Oct-89          809           162
Nov-89          837           191
Dec-89          897           202
Jan-90          927           228
Feb-90          997           235
Mar-90          1038          262
Apr-90          1525          301
May-90          1580          323
Jun-90          1639          338
Jul-90          1727          408
Aug-90          1894          452
Sep-90          1988          485
Oct-90          2063          527
Nov-90          2125          571
Dec-90          2190          615
Jan-91          2338          688        10
Feb-91          2417          717        20
Mar-91          2501          757        37
Apr-91          2622          793       421
May-91          2763          882       454
Jun-91          2982          989       583
Jul-91          3086         1012       616
Aug-91          3258         1066       666
Sep-91          3389         1128       818
Oct-91          3556         1214       869
Nov-91          3751         1302       907
Dec-91          4305         1450       948
Jan-91          4526         1496      1160

... snip ...

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Happy DEC-10 Day

Refed: **, - **, - **, - **, - **, - **
From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Happy DEC-10 Day
Newsgroups: alt.sys.pdp10, alt.folklore.computers
Date: Sun, 17 Jan 2010 10:47:19 -0500
jmfbahciv <jmfbahciv@aol> writes:
Would .11 seconds be "fast" enough for a typist? There's nothing more upsetting w.r.t. WPM than a very slight delay.

re:
http://www.garlic.com/~lynn/2010b.html#31 Happy DEC-10 Day
http://www.garlic.com/~lynn/2010b.html#32 Happy DEC-10 Day

mechanical typewriters ... had more than that delay on carriage return.

3270 had two modes for input ... no delay for whole screen of input (80x24, potentially 1920 chars) ... and enter ... in this case .11 seconds.

there was study of human interactions at research institution in early 70s ... that there was some variation in humans being able to differentiate between instantaneous response and some delay ... with the threshold ranging between .10 seconds .25 seconds.

later there was brain scan studies that there showed variation in different people in the time that signals propagated thru the brain (which possibly correlated with threshold of differentiating instantaneous and delay).

there was also a subsecond time-sharing study that consistent predictable delay ... say always 1second ... was better than random varying delay ... say varied between .5 second and 2 seconds.

the issue with the previously mentioned FIFO box hack for 3277 (hack wasn't possible for 3278 because so much of the electronics had been moved back into the controller) ... was that between the "enter key" being pushed (say on full-screen of input) and the screen update ... there would normally be a period when the keyboard wasn't taking input (keyboard would lock up if a key was hit and then the person would have to stop and reset the keyboard) ... the FIFO box hack allowed the person just to keep typing w/o having to worry about keyboard locking up (after 3270 enter was hit).

100wpm at 5chars/word ... is 600cpm (including space/wrod) or 10cps ... or about .10secs/char. so a fast typist would be ready to type the next char. before 3278 was ready. however 1) 3270 enter key is different akin to mechanical typewritter mechanical return ... and it would be sufficient that the delay is predictable and 2) the 3277 FIFO box would somewhat mask that many 3270 operations were really half-duplex.

(1982) "The Economic Value of Rapid Response Time"
http://www.vm.ibm.com/devpages/JELLIOTT/evrrt.html

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Happy DEC-10 Day

Refed: **, - **, - **, - **, - **, - **
From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Happy DEC-10 Day
Newsgroups: alt.sys.pdp10, alt.folklore.computers
Date: Sun, 17 Jan 2010 11:12:34 -0500
another item ..

[netinfo/gosip-order-info.txt]                             [ 9/91]

This information was compiled and made available by the National
Institute of Standards and Technology (NIST).

                                                        August, 1991

... some snipping

GOSIP Version 1.
----------------

GOSIP Version 1 (Federal Information Processing Standard 146) was published in
August 1988.  It became mandatory in applicable federal procurements in August
1990.

Addenda to Version 1 of GOSIP have been published in the Federal Register
and are included in Version 2 of GOSIP.  Users should obtain
Version 2.

GOSIP Version 2.
----------------

Version 2 became a Federal Information Processing Standard (FIPS) on
April 3, 1991 and will be mandatory in federal procurements initiated
eighteen months after that date, for the new functionality contained
in Version 2.  The Version 1 mandate continues to be in effect.
Version 2 of GOSIP supersedes Version 1 of GOSIP.  Version 2
of GOSIP makes clear what protocols apply to the GOSIP Version 1 mandate
and what protocols are new for Version 2.

... snip ...

The following "List of domains generated by Internet Domain Survey progam, October 1990" had over 9340 domains

including things like:


z1.fidonet.org
z1.ieee.org
z2.fidonet.org
z2.ieee.org
z3.fidonet.org
z3.ieee.org
z4.fidonet.org
z4.ieee.org
z5.fidonet.org
z5.ieee.org
z6.ieee.org
z7.ieee.org
z8.ieee.org
z89.ieee.org
z9.ieee.org
z99.ieee.org

... snip ...

then there is:


host: ftp.nisc.sri.com
directory: netinfo
file: internet-access-providers-non-us.txt
date: December 1992

... snip ... including entry for EARN:

6.1.1.2. EARN We would like to acknowledge and thank Nadine Grange of the EARN Office in France for the following information.

EARN, the European Academic Research Network, is the first general purpose computer network dedicated to universities and research institutions throughout Europe, the Middle East and Africa.

The network is widely used for scientific, educational, academic and research purposes. Commercial and political use is not allowed, either directly or indirectly.

EARN is made up of nearly 500 institutions including universities, European research centers (e.g., CERN, the European Space Agency, and the European Molecular Biology Laboratory), and national research centers and laboratories such as CNRS (France); Rutherford Appleton Laboratory (UK); CNR, INFN, and CINECA (Italy); DESY, GSI, DFLVR and the Max Planck Institute (Germany).

EARN also has links to 27 countries including Yugoslavia, Turkey, Algeria, Morocco, Tunisia, and Egypt, Iceland, and Luxembourg, to name a few.

EARN is an integral part of BITNET (see Section 1.5.4), in that it is based on the same protocols and shares the same name space. Through BITNET, EARN members have access to equivalent facilities in Argentina, Brazil, Canada, Chile, Japan, South Korea, Mexico, Singapore, Taiwan and the United States.

Most of the academic networks in the world can be accessed through EARN including EUnet, HEPnet, NSFNET, national European networks such as DFN in Germany and JANET in the UK, as well as a regional European Network such as NORDUnet, which links all the Nordic countries (see Section 6.28).

One of EARN's major objectives is to stimulate cooperative research, support the day-to-day exchange of research information, and the execution of joint projects and publications. Like BITNET, EARN supports mail, mailing lists, and a type of file transfer. It provides the LISTSERV mailing list function. Its facilities also allow users access to remote applications, databases, and libraries.

EARN is also an international member of RARE (Reseaux Associes pour la Recherche Europeenne) and cooperates actively with RARE and COSINE (Cooperation for Open Systems Interconnection Networking in Europe) on OSI for the research community. RARE and COSINE are more fully described in Sections 10.1.5 and 10.1.7.

For information about access to EARN, how to become a member organization or member country, or any other general information, contact your country's EARN representative or:

European Academic Research Network
BP 167
F-91403 Orsay CEDEX
FRANCE
BITNET/EARN/NetNorth: grange@frors12
Internet: grange%frors12.bitnet@mitvma.mit.edu
+33 1 69 82 39 73
FAX: +33 1 69 28 52 73


... snip ...

lots of past posts mentioning BITNET &/or EARN:
http://www.garlic.com/~lynn/subnetwork.html#bitnet

old email from person setting up EARN ... he had previously done a stint at the cambridge science center ... the following year we exchanged teenage offspring for the summer:
http://www.garlic.com/~lynn/2001h.html#email840320

historical reference regarding EARN and listserv:
http://www.lsoft.com/products/listserv-history.asp

listserv is similar to earlier TOOLSRUN application used on the internal network ... that offerred both a USENET kind of operation as well as mailing list operation.

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Happy DEC-10 Day

Refed: **, - **, - **
From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Happy DEC-10 Day
Newsgroups: alt.sys.pdp10, alt.folklore.computers
Date: Sun, 17 Jan 2010 13:29:42 -0500
"Charlie Gibbs" <cgibbs@kltpzyxm.invalid> writes:
Although I worked in Univac's OS/3 group, I heard a story from over the fence in the 1100 group that their approach to providing consistent response time was to delay responses that were ready too soon. I suppose that's easier than making slow responses faster, but it still didn't sound like Doing the Right Thing.

re:
http://www.garlic.com/~lynn/2010b.html#31 Happy DEC-10 Day
http://www.garlic.com/~lynn/2010b.html#32 Happy DEC-10 Day
http://www.garlic.com/~lynn/2010b.html#35 Happy DEC-10 Day

I heard jokes about TSO (batch system "time sharing option") attempting to do something similar ... choose a long enuf period that it was fairly easy to make ... and then make it fixed.

i had done dynamic adaptive resource management in the 60s as undergraduate ... sometimes referred to as the (wheeler) "fair share" scheduler ... since the default resource allocation was "fair share".
http://www.garlic.com/~lynn/subtopic.html#fairshare

however, the batch system didn't get anything like dynamic adaptive resource management until much more recently.

at one point in the early 80s ... there was a corporate stategic statement that CMS would be the official interactive platform ... and the TSO group contacted me about possibly considering rewriting the MVS resource manager (however, TSO has had a lot more performance issues than just the MVS resource manager). related old email
http://www.garlic.com/~lynn/2006b.html#email800310

things had slightly recovered from the mid-70s when in the wake of demise of future system effort ... the head of POK had convinced the corporation to kill vm370, shutdown the burlington mall vm370 group and transfer all the people to POK ... part of mad rush to get stuff back into the 370 product pipeline ... supposedly the vm370/cms people were needed to make the MVS/XA ship schedule, which still wasn't until the later part of the early 80s.

Endicott did manage to resurrect the vm370 product mission ... but had to reconsitute the development group from scratch. There is joke that head of POK was major contributor to VMS ... since so many of the burlington mall group weren't going to leave the area and went to work for DEC (also PRIME and some number of other companies in the greater boston area).

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Happy DEC-10 Day

Refed: **, - **, - **, - **, - **, - **, - **, - **, - **, - **
From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Happy DEC-10 Day
Newsgroups: alt.sys.pdp10, alt.folklore.computers
Date: Sun, 17 Jan 2010 13:40:56 -0500
Gene Wirchenko <genew@ocis.net> writes:
"Wah! Lynn is compentent, and we aren't!"?

re:
http://www.garlic.com/~lynn/2010b.html#31 Happy DEC-10 Day

but they would constantly take their revenge.

a couple recent posts mentioning business ethics is oxymoron and/or they will forgive you for being wrong, but they will never forgive you for being right:
http://www.garlic.com/~lynn/2009.html#53 CROOKS and NANNIES: what would Boyd do?
http://www.garlic.com/~lynn/2009e.html#37 How do you see ethics playing a role in your organizations current or past?
http://www.garlic.com/~lynn/2009h.html#74 My Vintage Dream PC
http://www.garlic.com/~lynn/2009o.html#36 U.S. students behind in math, science, analysis says
http://www.garlic.com/~lynn/2009p.html#87 IBM driving mainframe systems programmers into the ground
http://www.garlic.com/~lynn/2009r.html#50 "Portable" data centers

other recent posts mentioning being on the receiving end of some internal organization's wrath for one reason or another:
http://www.garlic.com/~lynn/2009h.html#68 My Vintage Dream PC
http://www.garlic.com/~lynn/2009h.html#71 My Vintage Dream PC
http://www.garlic.com/~lynn/2009k.html#17 Bulletproof
http://www.garlic.com/~lynn/2009k.html#49 A Complete History Of Mainframe Computing
http://www.garlic.com/~lynn/2009k.html#62 Hercules; more information requested
http://www.garlic.com/~lynn/2009o.html#17 Broken hardware was Re: Broken Brancher
http://www.garlic.com/~lynn/2009o.html#31 Justice Department probing allegations of abuse by IBM in mainframe computer market
http://www.garlic.com/~lynn/2009p.html#0 big iron mainframe vs. x86 servers
http://www.garlic.com/~lynn/2009q.html#74 Now is time for banks to replace core system according to Accenture

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Happy DEC-10 Day

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Happy DEC-10 Day
Newsgroups: alt.sys.pdp10, alt.folklore.computers
Date: Sun, 17 Jan 2010 14:14:43 -0500
Anne & Lynn Wheeler <lynn@garlic.com> writes:
in jan '83 when arpanet was making transition to internet (having approx. 100 nodes or 255 hosts ... since arpanet counted IMP network nodes as separate from the number of attached host processors) .. the internal network was getting close to passing 1000 nodes/hosts ... lots of the rapid growth being 43xx machines ... misc. past email mentioning 43xx
http://www.garlic.com/~lynn/lhwemail.html#43xx


old post with list of corporate locations around the world that added one or more new network nodes during 1983 (118 locations, from Amsterdam to Zurich and Vancouver to Johannesburg)
http://www.garlic.com/~lynn/2006k.html#8 Arpa address

one of the issues with corporate links was that they were required to be encrypted (if they left corporate premise). in '85 time-frame there was comment that the internal network had over half of all the link encryptors in the world ... past posts mentioning internal network
http://www.garlic.com/~lynn/subnetwork.html#internalnet

in the 70s & 80s there was enormous difficulty getting gov. approvals for deploying encrypted links ... especially when they crossed national boundaries.

encryption back then just about required some sort of embedded hardware support ... especially for higher speed links like we were doing in HSDT project
http://www.garlic.com/~lynn/subnetwork.html#hsdt

with T1 and higher speed links. old email mentioning 3081 processor doing software DES at 150kbytes/cpu second (would have needed dedicated two processor 3081K to handle encryption on each end of full-duplex T1 link)
http://www.garlic.com/~lynn/2006n.html#email841115

for other drift, other old email discussing a PGP-like public key infrastructure
http://www.garlic.com/~lynn/2007d.html#email810506
http://www.garlic.com/~lynn/2006w.html#email810515

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Daylight Savings Time again

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Daylight Savings Time again
Newsgroups: alt.folklore.computers
Date: Sun, 17 Jan 2010 14:46:28 -0500
greymaus <greymausg@mail.com> writes:
Kennedy had continued Eisenhower's policy of 'advising', and had sent in large numbers of 'advisors'. (These men, if asked, should have warned of what the real situation in Vietnam was. Probably, nobody did.)

there is at least one book that claims special forces had things relatively stable ... but westmoreland came in and wanted battles to give regular army officers field command experience (along with promotions)

quick search engine reference:
http://www.answers.com/topic/william-westmoreland

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Happy DEC-10 Day

Refed: **, - **, - **, - **
From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Happy DEC-10 Day
Newsgroups: alt.sys.pdp10, alt.folklore.computers
Date: Sun, 17 Jan 2010 15:06:56 -0500
jmfbahciv <jmfbahciv@aol> writes:
There were customers which did forms computing. BCTel was a TOPS-10 site which ran MCS. Responding to every character was eliminated unless the user was monitor mode.

JMF and CDO wrote a corporate architectural specification w.r.t. this kind of computing. IIRC, the corporate goal at that time was to be able to do 1000 transactions/second; the cybercrud assigned to this project was TPS.


re:
http://www.garlic.com/~lynn/2010.html#19 Happy DEC-10 Day
http://www.garlic.com/~lynn/2010.html#40 Happy DEC-10 Day
http://www.garlic.com/~lynn/2010.html#52 locate mode, was Happy DEC-10 Day
http://www.garlic.com/~lynn/2010.html#53 locate mode, was Happy DEC-10 Day
http://www.garlic.com/~lynn/2010b.html#32 Happy DEC-10 Day

references to Jim formalizing transaction definition
http://www.garlic.com/~lynn/2008p.html#27 Father Of Financial Dataprocessing

and instrumental in forming TPC
http://www.tpc.org/information/who/gray.asp

old email by Jim looking at ACP/TPF locking compared to system/r (120 transaction/second system)
http://www.garlic.com/~lynn/2008i.html#email800325
in this post
http://www.garlic.com/~lynn/2008i.html#39 American Airlines

and following post
http://www.garlic.com/~lynn/2008i.html#40 A Tribute to Jim Gray: Sometimes Nice Guys Do Finish First

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Happy DEC-10 Day

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Happy DEC-10 Day
Newsgroups: alt.sys.pdp10, alt.folklore.computers
Date: Sun, 17 Jan 2010 15:20:09 -0500
Anne & Lynn Wheeler <lynn@garlic.com> writes:
old email from person setting up EARN ... he had previously done a stint at the cambridge science center ... the following year we exchanged teenage offspring for the summer:
http://www.garlic.com/~lynn/2001h.html#email840320


re:
http://www.garlic.com/~lynn/2010b.html#36 Happy DEC-10 Day

major reason for the above email was I had gotten blamed for computer conferencing on the internal network in the late 70s and early 80s. misc. past posts mentioning internal network
http://www.garlic.com/~lynn/subnetwork.html#internalnet

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Daylight Savings Time again

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Daylight Savings Time again
Newsgroups: alt.folklore.computers
Date: Sun, 17 Jan 2010 18:54:29 -0500
re:
http://www.garlic.com/~lynn/2010b.html#40 Daylight Savings Time again
http://www.garlic.com/~lynn/2010b.html#42 Daylight Savings Time again

this article touches on it a little more
http://www.vietnamgear.com/bio/1.aspx

from above ..
Following a series of heavy ARVN defeats in May and June 1965, Westmoreland believed the Viet Cong were moving into the third and final phase of the insurgency - the fielding of large conventional style units. Consequently, rather pursuing a counterinsurgency approach based on population security he designed a strategy of attrition,

... snip ...

i.e. special forces trained to work with local populations.

I don't have the book I original saw the reference ... but was history of special forces (why was original organization named "10th group"?) ... I've seen copies in major bookstores in war history section.

further search engine turned up quotes from various "google books" (about westmoreland changing from special forces counterinsurgency to traditional army operations):
The Dynamics Of Defeat: The Vietnam War In Hau Nghia Province The Army and Vietnam

another reference here (discord between kennedy pushing special forces counterinsurgency and the mainstream army approach with conventional warfare):
http://chnm.gmu.edu/courses/schrag/wiki/index.php?title=The_Army_and_Vietnam

this was somewhat repeated in the stories about Boyd's battle plan for desert storm in contrast to traditional army conventional warfare approach ... I had sponsored Boyd's briefings at IBM ... some past posts mentioning Boyd &/or OODA-loops
http://www.garlic.com/~lynn/subboyd.html

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

sysout using machine control instead of ANSI control

Refed: **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **
From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: sysout using machine control instead of ANSI control
Newsgroups: bit.listserv.ibm-main, alt.folklore.computers
Date: Sun, 17 Jan 2010 19:48:59 -0500
chrismason@BELGACOM.NET (Chris Mason) writes:
[2] http://www.garlic.com/~lynn/gcard.html

[3] There is actually a mistake on the supposedly "green card" text in that the machine "Immediate" "space 0 lines" should be X'03', that is, a "no-operation" - because - think about it - nothing happens!


my gcard.html ... was q&d conversion of (CMS) IOS3270 file that was widely available internally, I wasn't the original author, but had added the *sense* information in the file ... most of which came from the 360/67 (blue) "reference card"; see bottom of web page for original attribution.

scan of "real" green card:
http://www.bitsavers.org/pdf/ibm/360/referenceCard/

gimick that VNET/RSCS used for networking control ("TAG") information was a no-op ('03') ccw pointing to the control information (with length of the information) ... aka VNET/RSCS used the cp (unit record) spool system for all its files ... so everything had to look like a printer or punch stream (generated with appropriate channel commands, cp treated the no-op as no-op ... but would still copy the indicated data into its spool file).

A major internal email client was VMSG. The PROFS group at one point acquired the source for an early copy of VMSG for the PROFS email client code. Later the VMSG author offerred to upgrade PROFS with latest VMSG version that had a lot more function ... the PROFS group denied that it was using VMSG and then attempted to have the VMSG author fired. That was suspended when the VMSG author pointed out that every PROFS message carried his initials in the comment portion of the RSCS network control ("TAG") information.

RSCS/VNET was the dominant internal networking infrastructure ... part of it was because of the NJE (HASP/JES) heritage using the HASP psuedo device table to define networking nodes. The internal network was larger than the arpanet/internet from just about the beginning until sometime mid-85 or early 86. misc. past posts mentioning internal network
http://www.garlic.com/~lynn/subnetwork.html#internalnet

The early HASP networking source code changes had "TUCC" out in cols. 68-71 ... and used the left-over entries in the one-byte index (255 entry) psuedo device table; installations frequently had 60-80 psuedo (unit record) devices (printer, punch, reader) ... which left less than 200 for defining networking nodes. By the time JES2/NJE shipped, the internal network was already over 200 nodes.

Corporate wasn't even going to allow RSCS/VNET to be announced (this was in the period when POK had convinced corporate to kill off vm370, shutdown the burlington mall development group and move all the people to POK ... justification was because POK needed all the people in order to meet the MVS/XA ship schedule; endicott eventually managed to save the vm370 product mission, but had to reconsitute a group from scratch; head of POK is also considered a major contributor to VAX/VMS because so many people left for DEC rather than move to POK).

The JES2/NJE group did manage to talk the corporation into a joint JES2/RSCS product announcement. The issue was that even at the minimum monthly rate that could be charged for corporate product ... it would still cover the VNET/RSCS development costs. However, there was no customer forecast at any monthly price that would cover the NJE development costs (number of customers times monthly rate was always less than NJE costs; as projected monthly rate went up, the number of forecasted customers declined). The way out for JES2 was to combine the RSCS+NJE costs divided by the combined RSCS+NJE customer forecast ... which finally resulted in number the business people could agree with.

VNET/RSCS had a fairly clean layered implementation ... which among other things allowed native drivers to co-exist with NJE drivers. In fact, VNET/RSCS quickly became mechanism to keep different JES2s from crashing MVS. JES2/NJE had jumbled up networking information and JES2 control information ... and network traffic between JES2 at different releases could result in JES2 failure, also taking down the MVS system.

As a result, there was a growing library of VNET/RSCS NJE drivers ... allowing the specific NJE driver for the release of JES2 on the other end of the link. The VNET/RSCS NJE drivers also had a growing body of code that would rewrite NJE headers (originating from another JES2 system) to be compatible with the directly connected JES2 system. There is the infamous case of a modified San Jose JES2 system causing MVS systems in Hursley to crash ... and it being blamed on VNET/RSCS (because the Hursley VNET/RSCS NJE drivers hadn't been updated with the latest countermeasures for keeping incompatible releases of JES2 causing MVS to crash). misc. past posts mentioning HASP, JES2, and/or hasp/jes2 networking
http://www.garlic.com/~lynn/submain.html#hasp

Native VNET/RSCS drivers continued to be used on the internal network long after the decision was made to only ship (non-native) NJE drivers ... in part, because the native VNET/RSCS drivers were significantly more efficient and got higher sustained throughput.

Similar VNET/RSCS technology (w/NJE drivers but w/o native drivers) was used during the 80s for educational BITNET in US and EARN in europe ... misc. past posts mentioning BITNET/EARN
http://www.garlic.com/~lynn/subnetwork.html#bitnet

By the time that JES2 got around to supporting 999 nodes, the internal network had exceeded 1000 nodes ... and by the time JES2 was upgraded to support 1999 nodes, the internal network had exceeded 2000 nodes. JES2 also had a nasty habit of not only discarding network traffic when the destination node wasn't defined in its table ... but would also discard traffic when the origin node wasn't in its table (so you never wanted JES2 in any critical location in the internal network)

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

sysout using machine control instead of ANSI control

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: sysout using machine control instead of ANSI control
Newsgroups: bit.listserv.ibm-main, alt.folklore.computers
Date: Mon, 18 Jan 2010 10:29:13 -0500
Anne & Lynn Wheeler <lynn@garlic.com> writes:
my gcard.html ... was q&d conversion of (CMS) IOS3270 file that was widely available internally, I wasn't the original author, but had added the *sense* information in the file ... most of which came from the 360/67 (blue) "reference card"; see bottom of web page for original attribution.

re:
http://www.garlic.com/~lynn/2010b.html#44 sysout using machine control instead of ANSI control

360/67 blue "reference card" was filled in with lots of sense info for various devices ... except for A220 (HYPERChannel link adapter) which I added ... I was using lots of HYPERChannel boxes in various HSDT activities
http://www.garlic.com/~lynn/subnetwork.html#hsdt

This particular 360/67 "reference card" is stamped with "M"s name (I must have borrowed and never returned). GML had been invented in 1969 by "G", "M", and "L" (first letter of their last names) at the science center ... which later was standardized as SGML ... misc. past posts
http://www.garlic.com/~lynn/submain.html#sgml

SGML then morphed into HTML, XML, etc ... old description of the evolution:
http://infomesh.net/html/history/early/

and then the first webserver outside cern is slac/vm system
http://www.slac.stanford.edu/history/earlyweb/history.shtml

... cern & slac were sister institutions ... sharing a lot of technology and applications ... both were on BITNET/EARN (educational network using similar RSCS technology used for internal network)

long-winded recent post with old early 90s information from netinfo center ... including long extract about EARN
http://www.garlic.com/~lynn/2010b.html#36 Happy DEC-10 Day

I've lots of bits and pieces from such places from the late 80s & early 90s. I had bunch of CMS execs that I would periodically execute which would go out to various places on the internet and check for new/change files ... and then pull off new/changed stuff. other references to archive
http://www.garlic.com/~lynn/2010b.html#23 Happy DEC-10 Day
http://www.garlic.com/~lynn/2010b.html#33 Happy DEC-10 Day
http://www.garlic.com/~lynn/2010b.html#34 Happy DEC-10 Day

for other drift ... old email related to nsfnet backbone
http://www.garlic.com/~lynn/lhwemail.html#nsfnet

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Oldest Instruction Set still in daily use?

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Oldest Instruction Set still in daily use?
Newsgroups: alt.folklore.computers
Date: Mon, 18 Jan 2010 11:19:46 -0500
a frequent refrain in news discussions looking at financial mess and why hasn't congress been able to take any action ... is that the financial industry "owns" congress.

a recent reference
http://www.garlic.com/~lynn/2010.html#81 Happy DEC-10 Day

CNN Fareed Zakaria had roundtable that discussed some of it, broadcast yesterday; the freakonomist author was part of the roundtable ... partial reference:
http://www.cnn.com/2010/OPINION/01/14/zakaria.wall.street.bonuses/index.html

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Daylight Savings Time again

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Daylight Savings Time again
Newsgroups: alt.folklore.computers
Date: Mon, 18 Jan 2010 11:30:40 -0500
jmfbahciv <jmfbahciv@aol> writes:
Note that the existence of atom bombs is also being ignored; just as it is being ignored today.

recent news item

Obama's Nuclear Arms Pledge Hits Stumbling Block
http://www.time.com/time/world/article/0,8599,1951850,00.html

misc. posts from last year mentioning DTRA:
http://www.garlic.com/~lynn/2009.html#57 CROOKS and NANNIES: what would Boyd do?
http://www.garlic.com/~lynn/2009.html#71 CROOKS and NANNIES: what would Boyd do?
http://www.garlic.com/~lynn/2009s.html#45 Audits VII: the future of the Audit is in your hands

DTRA website:
http://www.dtra.mil/

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Happy DEC-10 Day

Refed: **, - **, - **, - **, - **, - **, - **
From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Happy DEC-10 Day
Newsgroups: alt.sys.pdp10, alt.folklore.computers
Date: Mon, 18 Jan 2010 11:51:15 -0500
jmfbahciv <jmfbahciv@aol> writes:
1982? It took a long time to study that.

re:
http://www.garlic.com/~lynn/2010b.html#31 Happy DEC-10 Day
http://www.garlic.com/~lynn/2010b.html#35 Happy DEC-10 Day

it has been studied quite a bit by then ... however, this particular study was attempting to counteract a bunch of (early 80s) stuff from (MVS) TSO and 3274/3278 camps ... that subsecond response didn't provide any benefit and wasn't needed.

Lots of 3278 use was online (MVS) forms operations (IMS, CICS, etc) ... involving things like transcribing pieces of insurance form; hitting enter ... which then resulted in dbms update. The enter+dbms transaction delay was going to be more than small fraction of a second. It was different than bulk data entry (potentially just being accumulated in non-DBMS file).

However, TSO & 3274/3278 camps was attempting to then extend that scenario to interactive computing as justification for not needing subsecond response for anything.

Lots of the forms entry was also DBMS query operations ... current analog is lots of web-based browser forms.

my (partial) response to mask the current web delays ... is to queue up large number of web pages in different browser tabs ... and then browse the different tabs at local PC response ... rather than having to sychronously wait for each individual web page.

misc past posts mentioning sometimes having several hundred (browser) tabs:
http://www.garlic.com/~lynn/2005e.html#48 Mozilla v Firefox
http://www.garlic.com/~lynn/2005e.html#50 Mozilla v Firefox
http://www.garlic.com/~lynn/2005n.html#8 big endian vs. little endian, why?
http://www.garlic.com/~lynn/2005n.html#15 1.8b2 / 1.7.11 tab performance
http://www.garlic.com/~lynn/2005n.html#41 Moz 1.8 performance dramatically improved
http://www.garlic.com/~lynn/2005o.html#13 RFC 2616 change proposal to increase speed
http://www.garlic.com/~lynn/2007l.html#30 tab browsing
http://www.garlic.com/~lynn/2007m.html#8 John W. Backus, 82, Fortran developer, dies
http://www.garlic.com/~lynn/2008b.html#32 Tap and faucet and spellcheckers
http://www.garlic.com/~lynn/2008b.html#35 Tap and faucet and spellcheckers
http://www.garlic.com/~lynn/2008h.html#71 Mainframe programming vs the Web
http://www.garlic.com/~lynn/2008i.html#85 Which of the latest browsers do you prefer and why?
http://www.garlic.com/~lynn/2009q.html#72 Now is time for banks to replace core system according to Accenture

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

search engine history, was Happy DEC-10 Day

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: search engine history, was Happy DEC-10 Day
Newsgroups: alt.sys.pdp10, alt.folklore.computers
Date: Mon, 18 Jan 2010 14:15:01 -0500
Maarten van Tilburg <mtilburg@planet.nl> writes:
But SAP could not run on VMS (that was the word, I never knew if it was actually true, SAP refused to support it), so there we were: the finanial system was implemented on Windows. With at least one single point of failure (the database server), a backup system of the database which was unsecure (Oracle had no on-line backup like RDB) a needing three times as much people to support the system than the previous one.

we ran into some large corporations that got into the SAP camp ... and found themselves spending $50m/annum on SAP consultants.

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Happy DEC-10 Day

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Happy DEC-10 Day
Newsgroups: alt.sys.pdp10, alt.folklore.computers
Date: Mon, 18 Jan 2010 17:49:53 -0500
Charles Richmond <frizzle@tx.rr.com> writes:
Not to forget the APL character set. At the college I attended, the printing APL terminals were all DECWriters. The characters produced did *not* work well for "spirit duplicator" style copies. And the Math and CS departments did *not* have a 2741 typeball. All our APL tests were hand written. :-(

about the only thing I have left of 2741 I had at home (can't even find any pictures) is the apl typeball ... picture here:
http://www.garlic.com/~lynn/lhwemail.html#oldpicts

In the 80s ... I wanted a tool case ... and put in order for standard FE tool case. I got a lot of push back because I wasn't part of field service ... but eventually was able to push the order through (looks like expensive large leather covered briefcase). Lots of the tools in it appear to related to doing maintenance on selectric typewriters.

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Source code for s/360

Refed: **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **
From: lynn@GARLIC.COM (Anne & Lynn Wheeler)
Subject: Re: Source code for s/360
Newsgroups: bit.listserv.ibm-main
Date: 18 Jan 2010 14:43:42 -0800
lindy.mayfield@SSF.SAS.COM (Lindy Mayfield) writes:
i can do that on vm? create my own instruction?

originally virtual machine system was cp40 done on a 360/40 with custom hardware modifications to support virtual memory.

when standard virtual memory became available with 360/67 ... cp40 morphed into cp67.

370 was originally announced pretty much the same as 360 ... with a few new instructions, but w/o virtual memory.

there was a special project jointly between the science center and endicott to modify cp67, implementing 370 virtual machines (supporting the full set of unannounced 370 virtual memory features ... various bits and fields differed from 360 virtual memory architecture).

there was also a set of modifications to cp67 that would run with 370 virtual memory hardware (instead of 360/67 virtual memory). that was up and running in 370 virtual machine (on cp67 running on real 360/67) a year before the first engineering 370 with virtual memory hardware was operational (a 370/145 in endicott).

there was a security issue at the science center since they had some number of non-employee users of the cp67 system from various educational institutions in the boston area. so to help avoid unannounced 370 virtual memory info from leaking; the standard operation was:

real 360/67 hardware "cp67l" system w/o 370 modifications "cp67h" system running in 360/67 virtual machine providing 370 virtual machines "cp67i" system running in a 370 virtual machine providing 370 virtual machines "cms" running in 370 virtual machine

non-employees using the "cp67l" system wouldn't have visability into what the "cp67h" system was doing in a separate virtual machine (or that there was 370 virtual machines or "cp67i" systems).

when engineering 370s with virtual memory support became available, they were normally run with the "cp67i" system ... long before vm370 became available. Internally, there was also "cp67sj" system ... which was a "cp67i" system with modifications done by San Jose with device support for 3330 disks and 2305 fixed head paging devices.

After the 23jun69 "unbundling" announcement ... starting to charge for application software, SE services, and other stuff ... there was issue with training for new SEs. misc. past posts mentioning unbundling
http://www.garlic.com/~lynn/submain.html#unbundle

New SEs had previosly gotten a lot of experience ... essentially in an apprentice position as part of large SE team at customer sites. With starting to charge for SE serices ... nobody could figure out how to do the "apprentice" thing. As a result, several internal CP67 virtual machine datacenters were set up as part of "HONE" ... supposedly to give SEs in branch offices, remote/online "hands-on" experience running various operating systems in cp67 virtual machines.

After the initial 370 announcement, a subset of the "cp67h" changes were applied to the HONE systems ... to allow (non-virtual memory) 370 virtual machines (supporting the new instructions in the original 370 announcement). This would allow SEs to build & test operating systems for "370" operation. misc. past posts mentioning HONE:
http://www.garlic.com/~lynn/subtopic.html#hone

the science center had also ported apl\360 to cms for cms\apl. some number of sales & marketing support applications started to be implemented (in apl) and also deployed on HONE. eventually the sales & marketing (apl) applications became so extensive that they eventually completely crowded out the SE virtual operating system activity. At some point, branch office sales had to process customer orders thru various HONE applications before they could be submitted (and HONE datacenters started to pop up around the world). One of my hobbies was supporting HONE operation ... and as a new employee fresh out of college ... I got some number of overseas trips as part of the HONE proliferation.

I had maintained large amount of cp67 & cms source (replicated on multiple tapes) up through the mid-80s. however, at one point there was an operational problem at the Almaden datacenter where random tapes were being selected for mounting as scratch ... and all tapes were overwritten. Old email pulling bits & pieces from the tapes for Melinda Varian (not long before they were lost):
http://www.garlic.com/~lynn/2006w.html#email850906
http://www.garlic.com/~lynn/2006w.html#email850908
http://www.garlic.com/~lynn/2006w.html#email850908b
in these old posts
http://www.garlic.com/~lynn/2006w.html#42
http://www.garlic.com/~lynn/2006w.html#48

Melinda's webpage:
http://www.leeandmelindavarian.com/Melinda/

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

The Shannon limit

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: The Shannon limit
Newsgroups: alt.folklore.computers
Date: Tue, 19 Jan 2010 08:45:33 -0500
The Shannon limit
http://web.mit.edu/newsoffice/2010/explained-shannon-0115.html

from above:
A 1948 paper by Claude Shannon SM '37, PhD '40 created the field of information theory -- and set its research agenda for the next 50 years.

... snip ...

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Happy DEC-10 Day

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Happy DEC-10 Day
Newsgroups: alt.sys.pdp10, alt.folklore.computers
Date: Tue, 19 Jan 2010 09:11:37 -0500
jmfbahciv <jmfbahciv@aol> writes:
The subject was multi-CPU systems. The powers wanted some flavor of SMP and got master/slave instead because of his obdurate attitude.

old post with 8Mar88 (VMS 5/vax 8800) product announcement eventually getting around to SMP
http://www.garlic.com/~lynn/2007.html#46 How many 36-bit Unix ports in the old days?

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

ISPs could cut spam easily, says expert

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: ISPs could cut spam easily, says expert
Newsgroups: alt.folklore.computers
Date: Tue, 19 Jan 2010 09:19:58 -0500
ISPs could cut spam easily, says expert
http://news.techworld.com/security/3210489/isps-could-cut-spam-easily-says-expert/

I mentioned something similar after the greencard spam in the 90s. Issues raised why they wouldn't 1) they make money from the spammers, 2) they didn't want potential legal actions if they started blocking and something slipped through, 3) routers and servers used by most ISPs didn't have the processing capability
http://en.wikipedia.org/wiki/Newsgroup_spam

A counter to #3 was that ISPs were starting to block multiple concurrent connections for the same account (like multiple dialup connections) ... and if they could figure out that ... they could reasonably figure out a spamming profile to block

Not long later was on business trip to Scottsdale and was having dinner at a mexican restaurant in oldtown. A couple came in and was seated behind us and they were then joined by a man who sat behind me. The man spent an hr explaining how he did his spamming and how he could do it for their commercial website ... and some advice about their server configuration to ignore any irate responses to the spam.

past posts mentioning that dinner in scottsdale:
http://www.garlic.com/~lynn/2002o.html#34 Follklore
http://www.garlic.com/~lynn/2003e.html#61 hee-hee. I can do something about this spam
http://www.garlic.com/~lynn/2005i.html#14 The Worth of Verisign's Brand
http://www.garlic.com/~lynn/2009p.html#15 Rogue security software threat will grow in 2010, warns report

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Oldest Instruction Set still in daily use?

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Oldest Instruction Set still in daily use?
Newsgroups: alt.folklore.computers
Date: Tue, 19 Jan 2010 10:45:31 -0500
re:
http://www.garlic.com/~lynn/2010b.html#46 Oldest Instruction Set still in daily use?

related blog entry:

The Baseline Scenario

What happened to the global economy and what we can do about it

A Trap Of Their Own Design
http://baselinescenario.com/2010/01/19/a-trap-of-their-own-design/

about Baseline Scenario
http://baselinescenario.com/about/

also from the above: Financial Crisis for Beginners
http://baselinescenario.com/financial-crisis-for-beginners/

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Happy DEC-10 Day

Refed: **, - **, - **, - **, - **, - **, - **, - **
From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Happy DEC-10 Day
Newsgroups: alt.sys.pdp10, alt.folklore.computers
Date: Tue, 19 Jan 2010 10:55:33 -0500
Walter Bushell <proto@panix.com> writes:
And that will end Social Security. Once it's seen as a welfare program, it's death is certain whether by immediate cut or by neglect. Or just by understating the rate of inflation when computing benefits, as now.

re:
http://www.garlic.com/~lynn/2010b.html#19 STEM crisis
http://www.garlic.com/~lynn/2010b.html#24 Happy DEC-10 Day

nearly all the (pay-as-you-go) retirement benefit plans were based on the baby boomer generation being so much larger than the previous generation ... and their enormous wage earnings during their prime working years ... along with improved higher educational level and higher earning jobs (for that matter not just retirement benefits but nearly all the gov. funded operations dependent on tax revenues).

that is being inverted as the baby boomers retire ... and the following generation is only half the size and less well educated.

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Happy DEC-10 Day

Refed: **, - **, - **
From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Happy DEC-10 Day
Newsgroups: alt.sys.pdp10, alt.folklore.computers
Date: Tue, 19 Jan 2010 11:31:11 -0500
Walter Bushell <proto@panix.com> writes:
But the following generations need SS et. al. to get the Boomers to retire as the economy just cannot generate enough jobs.

re:
http://www.garlic.com/~lynn/2010b.html#19 STEM crisis
http://www.garlic.com/~lynn/2010b.html#24 Happy DEC-10 Day
http://www.garlic.com/~lynn/2010b.html#56 Happy DEC-10 Day

sort of chicken and egg ... only half as many people and less well educated ... contributed significantly to jobs moving out of the country.

at annual state governors convention in early 90s ... they looked at the falling education level and had a study that indicated if they could bring back the boomer STEM education levels ... it would add couple percent to GDP growth (along with more and higher paying jobs). STEM just kept falling rather than improving (along with the jobs).

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

How long for IBM System/360 architecture and its descendants?

Refed: **, - **, - **, - **, - **
From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: How long for IBM System/360 architecture and its descendants?
Newsgroups: alt.folklore.computers
Date: Tue, 19 Jan 2010 14:15:17 -0500
Eric Chomko <pne.chomko@comcast.net> writes:
But using the concept to write an OS, always written in assembly code in the old days, in a high-level language did come out before RISC. So, why not make a limited instruction set to force the users to go that route already in place. Bootstrap languages, high-level languages used to write OSs, and THEN RISC... Here we are today.

re:
http://www.garlic.com/~lynn/2010b.html#22 How long for IBM System/360 architecture and its descendants?

old email with little motherhood statements on the subject:
http://www.garlic.com/~lynn/2006t.html#email810812
in this post
http://www.garlic.com/~lynn/2006t.html#7 32 or even 64 registers for x86-64?

RISC ... reaction to (failed) future system ... go to the opposite extreme in hardware complexity ... and use pl.8 language and cp.r system to compensate for hardware short comings.

most of the (801/risc) Iliad chip projects got canceled for one reason or another ... resulting in some number of the engineers leaving and going to work for other vendors on next generation of risc efforts.

(801/risc) ROMP chip project (joint between research and office products) was pure pl.8 and cp.r ... for displaywriter follow-on. for various reasons that got canceled (one was the minimum ROMP displaywriter entry point ... price & performance; was above the top end of the existing displaywriter market place). they were then looking around and decided on using it for the growing unix workstation market.

prices of chips/hardware declined to the point where it became much less expensive to produce hardware for computers ... however, proprietary operating systems was still barrier. unix was emerging as an relatively inexpensive alternative. the ROMP group then hired the company that had done PC/IX ... to do aix v2. There was an issue with what to do with all the pl.8 programmers ... so they defined something called the VMR (implemented in pl.8) that provided a abstract virtual machine ... and a claim that it would take the PC/IX company much less time to port to the abstract virtual machine interface (than to the bare hardware).

This was somewhat disproved when the palo alto group did a port of BSD to the bare hardware. It was also a pain for things like new device drivers ... having to do both a pl.8 VRM device driver as well as a C unix device driver.

There were also some number of tweaks that had to do for ROMP for unix ... since original 801/ROMP didn't have things like protection/privilege domains. the claim was that pl.8 would only produce correct code ... and cp.r would only load correct pl.8 code for execution. moving to unix at least required hardware support separation between kernel and application programs.

the 801/risc virtual memory segment registers & inverted tables simplified (hardware) virtual memory operation. the issue was (at least for most of the 32bit address lifetime) 16 256mbyte segment registers made for a more difficult "sharing" model. the pl.8/cp.r response was that with no protection domains, the pl.8 application code would be able to switch segment register values as trivially as changing addresses in general purpose registers.

The change to unix model elimiante that possibility. At some point I got asked how to work out the details of mapping multiple different shared memory "objects" into single segment register for sharing ... old email on "romp small shared segment": reference:
http://www.garlic.com/~lynn/2006y.html#email841114c
http://www.garlic.com/~lynn/2006y.html#email841127
in this post
http://www.garlic.com/~lynn/2006y.html#36 Multiple mappings

misc. other old email mentioning 801, risc, iliad, romp, rios, etc
http://www.garlic.com/~lynn/lhwemail.html#801

another scenario was that there was no cache consistency ... so no multiprocessor ... also no cache consistency between i-cache and d-cache ... so things like loaders that would bring in programs and possibly operate on them as data as part of preparing for execution ... needed special instructions to force data from d-cache lines back to memory and potentially invalidate corresponding addresses in i-cache.

lots of old posts mentioning 801, risc, iliad, romp, rios, power, power/pc, etc.
http://www.garlic.com/~lynn/subtopic.html#801

some past references to John:
http://domino.research.ibm.com/comm/pr.nsf/pages/news.20020717_cocke.html
http://domino.watson.ibm.com/comm/pr.nsf/pages/bio.cocke.html
http://www.iment.com/maida/tv/computer/johncocke.htm
http://www.thocp.net/biographies/cocke_john.htm

John liked to go out drinking after work ... and good part of the 80s I lived in San Jose but worked for Yorktown ... so commuted from San Fran to Kennedy a couple times a month (work Monday in san jose and take redeye to kennedy and be at the office by 7am Tuesday). sometimes I would get shanghi'ed into going drinking with John Tuesday night (after only 4hrs sleep the night before) and he would want to stay out until wee hours of weds. morning.

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Happy DEC-10 Day

Refed: **, - **, - **, - **, - **, - **, - **, - **
From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Happy DEC-10 Day
Newsgroups: alt.sys.pdp10, alt.folklore.computers
Date: Tue, 19 Jan 2010 14:57:30 -0500
re:
http://www.garlic.com/~lynn/2010b.html#19 STEM crisis
http://www.garlic.com/~lynn/2010b.html#24 Happy DEC-10 Day
http://www.garlic.com/~lynn/2010b.html#56 Happy DEC-10 Day
http://www.garlic.com/~lynn/2010b.html#57 Happy DEC-10 Day

... part of STEM crisis

Government Finds U.S. Slipping In Tech Dominance; The U.S. lead in science, engineering, and technology is slipping as Asia's capabilities rise, report says.
http://www.informationweek.com/news/government/policy/showArticle.jhtml?articleID=222301424

part of the studies from the early 90s referred to possible tipping point ... because of declining domestic educational levels ... more and more foreign workers were imported to fill high-skill jobs ... but a critical mass of foreign workers could be reached ... where it would switch from importing the workers here .... to the jobs going to where they are (sort of analog to the concentration of skilled jobs in silicon valley from the last century ... but no longer in this country).

besides the retiring baby boomers increasing the ranks of retirees by a factor of four (and therefor increasing benefits payout by factor of four) ... and their replacements being only half as many and lower level of education (so ratio of productive workers being taxed to pay for the benefits ... to the number of retirees receiving benefits drops by factor of eight ... or generation following baby boomers has to work to support eight times as many retirees) ... the generation following baby boomers is going to account for drastically decreased consumer spending (half as many, as well as lower-skill, lower-paid jobs) ... decreased consumer spending will result in further loss of jobs ... which results in further decreases in consumer spending ... "negative feedback loop".

The transition to economy with much smaller ratio of workers to retirees as well as lower paid and lower skilled jobs ... is likely to be traumatic before things reach some stable economic level ... but likely with nearly everybody having much lower standard of living.

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Happy DEC-10 Day

Refed: **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **
From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Happy DEC-10 Day
Newsgroups: alt.sys.pdp10, alt.folklore.computers
Date: Tue, 19 Jan 2010 21:12:53 -0500
Anne & Lynn Wheeler <lynn@garlic.com> writes:
The transition to economy with much smaller ratio of workers to retirees as well as lower paid and lower skilled jobs ... is likely to be traumatic before things reach some stable economic level ... but likely with nearly everybody having much lower standard of living.

re:
http://www.garlic.com/~lynn/2010b.html#59 Happy DEC-10 Day

david walker (former comptroller general) was on last night's john stewart's daily show ... didn't see it last night ... but it was rebroadcast today a 7pm. walker was pitching his new book "comeback america".
http://www.amazon.com/Comeback-America-Turning-Restoring-Responsibility/dp/1400068606

walker made comment that fiscal responsibility bill expired in 2002 ... and it was after that ... was when he starting speaking out (comments like no congressmen for the last 50 yrs have been capable of middle school arithmetic). he made some reference that gov. debt is now $500,000 per person.

recent posts in thread mentioning walker:
http://www.garlic.com/~lynn/2010.html#36 Happy DEC-10 Day
http://www.garlic.com/~lynn/2010.html#37 Happy DEC-10 Day

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Source code for s/360 [PUBLIC]

Refed: **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **
From: lynn@GARLIC.COM (Anne & Lynn Wheeler)
Subject: Re: Source code for s/360 [PUBLIC]
Newsgroups: bit.listserv.ibm-main
Date: 19 Jan 2010 20:15:01 -0800
mike@MENTOR-SERVICES.COM (Mike Myers) writes:
On thinking it over some, it had to be early 1968, ending around March or so (I just reviewed my CV and found I went to FE education in April of 1968). So maybe it was closer to Release 11 or 12. Wasn't release 14 actually 14/15 (not that that means anything, just a recollection)?

re:
http://www.garlic.com/~lynn/2010b.html#51 Source code for s/360

part of presentation at '68 Atlantic City Share meeting on bunch of performance enhancements i had done at the univ to both mft14 & cp67. cp67 was installed at univ. where i was undergraduate ... and also os/360 support.
http://www.garlic.com/~lynn/94.html#18 CP/67 & OS MFT14

i had been doing highly customized os/360 sysgen's attempting to radically improve thruput ... as well as being able to do (at least) stage2 sysgen in production jobstream. I did lots of re-ordering stage2 sysgen to carefully place files and pds members on disk to optimize disk arm motion (i got about 3times thruput improvement for typical student fortran job stream ... this was before watfor and student jobs going thru standard fortgclg).

for cp67 ... i rewrote lots of kernel to significantly reduce pathlength.

os/360 releases i did for mft were 9.5, 11, and then 14. combined 15/16 came out ... and i did a mvt generation (mvt was starting to get to point that it was more reliable). big thing i remember about 15/16 was that it introduced format enhancement and being able to specify cylinder for vtoc (rather to defaulting to cylinder zero). I placed vtoc in middle of system packs ... and then attempted to force placements radiating out from the middle of the pack on both sides of the vtoc.

one of the problems were that typical PTF activity "replaced" PDS members ... and 5-6 months of system PTF activity could significantly degrade my carefully optimized thruput ... i.e. pds members being replaced in system datasets ... creating lots of gas ... standard pds compression didn't offer anyway of controlling member ordering. If I wasn't planning on doing near term sysgen for new system ... i would have to rebuild system to get back disk arm optimization.

end of '68, boeing was putting together basis of boeing computer service (BCS) ... moving datacenters from cost center to P&L basis ... at least on paper. As part of concept of "selling" services ... they wanted to add CP67 online timesharing ... and be able to sell CP67 internally within Boeing ... but also to external organizations (somewhat akin to some of the other commercial online cp67-based timesharing service bureaus that had been formed). Spring break, '69, they con'ed me into teaching one week class for the burgeoning BCS technical staff. Boeing then brought me in for summer ... they did some sort of paperwork that listed me as mid-level fulltime employee (that got me special parking lot privileges at boeing field). They brought in new 360/67 "simplex" ... installed in the hdqtrs machine room next to 360/30 that was doing payroll (deal was also cut that for the summer work, i also got some "special project" academic credit towards graduation).

The boeing huntsville two-processor 360/67 smp was also moved up to seattle that summer. It had been running as two single processors with modified version of mvt13. Boeing huntsville was supporting a lot of long running 2250 graphics applications. os/360 had significant problem with storage fragmentation with long running jobs. mvt13 had been modified to run with 360/67 virtual memory tables ... it didn't do any paging ... but used to re-org storage mapping to make it look contiguous (countermeasure to os/360 storage fragmentation).

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Happy DEC-10 Day

Refed: **, - **, - **, - **, - **, - **, - **
From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Happy DEC-10 Day
Newsgroups: alt.sys.pdp10, alt.folklore.computers
Date: Tue, 19 Jan 2010 23:56:55 -0500
Michael Wojcik <mwojcik@newsguy.com> writes:
Yep. The HTTP request/response model, plus the relatively high latency of the verbose HTTP protocol, plus the links-and-forms orientation of HTML, make traditional web apps very much block-mode.

The point of AJAX, of course (and crossing threads again), is to emulate a lower-latency interaction model on top of HTTP+HTML, by pushing the slow network I/O into the background and using client-side processing to provide a fancier front end. It's just the block-mode versus character-mode dichotomy again.


re:
http://www.garlic.com/~lynn/2010b.html#48 Happy DEC-10 Day

so i mentioned before doing this work on dbms, distributed lock manager, and cluster scaleup.
http://www.garlic.com/~lynn/2010b.html#32 Happy DEC-10 Day

well two of the people mentioned in this jan92 meeting ... referenced in this post
http://www.garlic.com/~lynn/95.html#13

later leave their posts and show up at a small client/server startup responsible for something called the "commerce server". by then we had also left ... in part because the cluster scaleup work had been transferred, announced as supercomputer (numerical intensive only ... no dbms stuff), and we had been told that we couldn't work on anything with more than four processors; ... misc. old email related to cluster scaleup
http://www.garlic.com/~lynn/lhwemail.html#medusa

so they wanted to do payment transactions on their server ... and the startup had invented this technology called "SSL" they wanted to use. We got this thing called "payment gateway" deployed ... using ha/cmp configuration (from the non-scaleup part of the cluster work) ... some past posts
http://www.garlic.com/~lynn/subnetwork.html#payments

there were various and sundry other things around the server edges for various security and internet attack resistance.

During this period ... there was rapid growth in both HTTP and HTTPS ... and both HTTP and HTTPS were seeing severe performance problems. HTTP/HTTPS were using TCP ... which had been designed/implemented for long-running sessions ... not for quick transactions. TCP had a minimum 7 packet exchange operation with relatively long tail in FINWAIT. High rate of HTTP activity and the TCP FINWAIT list exploded ... most implementations started finding that webservers were spending 95% of the processor running FINWAIT list. The small client/server startup had webservers for downloading their products ... and were adding servers almost as fast as they could be installed. Finally they installed a SEQUENT machine running Dynix ... and the problems cleared up ... SEQUENT had already fixed the long FINWAIT list issue in DYNIX to handle installations with 20,000 (real long running) telnet sessions. It took the other vendors another six months or so before there was new releases addressing the FINWAIT problem.

HTTPS shared the HTTP TCP problems ... but also had all its own crypto gorp ... digital certificate processing, encryption key exchange, encrypted data, etc ... way, over and above the minimum 7 packet tcp exchange.

So there is all the stuff with SSL digital certificates ... misc. posts
http://www.garlic.com/~lynn/subpubkey.html#sslcert

but there is also this DNSSEC stuff ... which SSL Certification Authority industry has somewhat been backing ... because it helps with the integrity of their certification processes for SSL digital certificates ... but it also represents a catch-22 for that industry
http://www.garlic.com/~lynn/subpubkey.html#catch22

one of the things that is part of DNSSEC is the ability to register public keys with the domain name authority ... and use the DNS infrastructure to do real time retrieval of public keys ... w/o the need for digital certificates.

now, recent post mentioning xtp technical advisory board and hsp
http://www.garlic.com/~lynn/2010b.html#4 Happy DEC-10 Day

... however, one of the features of XTP is a minimum of 3-packet exchange for reliable transmission (compared to the 7-packet minimum for tcp). If DNSSEC public keys were to be registered and clients could request any public keys to be piggy-backed on the response to domain name lookup (aka request to translate domain name to ip-address) ... combined with XTP ... it would be able to do a HTTPS-light in three-packet exchange w/o need for any of the digital certificate processing gorp.

the client gets the server's public key back in the same DNS response that it gets the server's ip-address. It then generates a random symmetric key ... encodes the transaction with the symmetric key ... and encodes the symmetric key with the server's public key. It then sends off the (XTP) transaction with the encoded symmetric key followed by the encoded transaction. The server gets the transactions, decodes the symmetric key with the server's private key ... and then decodes the transaction with the random symmetric key. The server then generates the response ... first encoding it with the client's random symmetric key and sends back the encrypted response. The client then decodes the response with the symmetric key that it had previously generated.

purely single round-trip ... with the same encryption strength of standard HTTPS ... but w/o all the extraneous round-trips and certificate overhead processing.

part of this was from responding to some payment protocol specification work in the mid-90s that was looking at an fully end-to-end payment protocol with appended (payment industry) digital certificates. However, the standard digital certificate payload is about 100 times larger than the base payment transaction payloads ... and add about 100 times in processing. misc. past posts discussing the enormous processing and payload bloat of some of these payment protocol specifications
http://www.garlic.com/~lynn/subpubkey.html#bloat

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Source code for s/360 [PUBLIC]

Refed: **, - **, - **, - **, - **, - **, - **, - **, - **, - **
From: lynn@GARLIC.COM (Anne & Lynn Wheeler)
Subject: Re: Source code for s/360 [PUBLIC]
Newsgroups: bit.listserv.ibm-main
Date: 20 Jan 2010 05:53:01 -0800
mike@MENTOR-SERVICES.COM (Mike Myers) writes:
I worked on a couple of projects in my two years there, CLEAR, which was a library source management system used to maintain OS/360 source code and the MEMMAP project mentioned earlier.

one of the problems that the JES2 group had was that they did all their source maintenance on CMS using CMS source maintenance ... recent JES2/HASP networking reference
http://www.garlic.com/~lynn/2010b.html#44 sysout using machine control instead of ANSI control

but then had to contort into CLEAR for release. This post refers to some old email with Melinda about providing early source to Melinda ...
http://www.garlic.com/~lynn/2010b.html#51 Source code for s/360

which was the source for the original implementation of the CMS multi-level source maintenance procedures
http://www.garlic.com/~lynn/2010b.html#51 Source code for s/360

originally done for the multi-level updates for "cp67l", "cp67h", and "cp67i" systems. The implementation/design was eventually directly supported by the editors and update program (rather as exec front-end processes) and eventually used for both cp67 and vm370 products ... and shipped as part of doing customer source level maintenance (i.e. fixes & updates were shipped as source updates).

there was later some folklore about some gov. agencies ... possibly mentioned here
http://web.archive.org/web/20090117083033/http://www.nsa.gov/research/selinux/list-archive/0409/8362.shtml

requesting the complete source that exactly corresponded to specific running MVS image. supposedly a corporate task force spent enormous amount of resources & money before coming back with conclusion that it wasn't practical.

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

SYSENTER/SYSEXIT_vs._SYSCALL/SYSRET

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: SYSENTER/SYSEXIT_vs._SYSCALL/SYSRET
Newsgroups: comp.arch
Date: Wed, 20 Jan 2010 09:09:43 -0500
Terje Mathisen <"terje.mathisen at tmsw.no"> writes:
Anything that updates a real memory location every us is a performance bug!

If you instead use a memory-mapped timer chip register, then you've still got the cost of a real bus transaction instead of a couple of core-local instructions.


one of the justification for the 370 timer facilities. 360s had location "80" timer in low-store. lower-end 360 modules updated in millisecond range ... higher end 360s updated low order bit every 13+ microseconds.

for compatibility, 370s did provide support for location 80 timer but at the millisecond range.

univ. where i was undergraduate had 360/67 (that had "high-speed" location 80 timer). I had been doing a bunch of enhancements to (virtual machine) cp67 ... one of which was adding tty/ascii terminal support to cp67. part of this was I attempted to do something with the 2702 terminal controller that it couldn't quite do (but should). somewhat as a result, the univ. started a clone controller project ... using an interdata/3, reverse engineer the 360 channel interface, build channel interface board for the interdata/3, program the interdata/3 to emulate 2702 controller with some additional function (later four of us got written up for being responsible for mainframe clone controller business).

some early controller tests resulted in bringing down the 360/67 (hardware "red-light"). the issue was the memory bus was shared between processor, the location 80 timer, and i/o channels (and these were non-cache machines). the location 80 timer had some leeway if the bus was in use when timer tic'ed ... but if the timer tic'ed again ... and there was previous timer memory update still pending ... the machine would stop/red-light.

had to go back and redo the controller channel board to make sure that it periodically told the channel to release the memory bus (in middle of transfers) so that any pending timer tic update could occur.

misc. past posts mentioning clone controller effort
http://www.garlic.com/~lynn/subtopic.html#360pcm

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Source code for s/360 [PUBLIC]

From: lynn@GARLIC.COM (Anne & Lynn Wheeler)
Subject: Re: Source code for s/360 [PUBLIC]
Newsgroups: bit.listserv.ibm-main
Date: 20 Jan 2010 06:32:56 -0800
gabe@GABEGOLD.COM (Gabe Goldberg) writes:
Here's the Web page I found, it's stories from Stretch/Harvest, a couple of notable S/360 predecessors:
http://users.bestweb.net/~collier/sh/stories.html -- with the obit I


re:
http://www.garlic.com/~lynn/2010b.html#51 Source code for s/360
http://www.garlic.com/~lynn/2010b.html#61 Source code for s/360 [PUBLIC]
http://www.garlic.com/~lynn/2010b.html#63 Source code for s/360 [PUBLIC]

other stretch/harvest at clemson.edu
http://www.cs.clemson.edu/~mark/stretch.html

at historical computer design page:
http://www.cs.clemson.edu/~mark/hist.html

also ACS
http://people.cs.clemson.edu/~mark/acs.html

and Future System section
http://people.cs.clemson.edu/~mark/fs.html

... above includes references to my periodic FS postings
http://www.garlic.com/~lynn/submain.html#futuresys

computer history museum harvest
http://www.computerhistory.org/collections/accession/102621818

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

SYSENTER/SYSEXIT_vs._SYSCALL/SYSRET

Refed: **, - **, - **
From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: SYSENTER/SYSEXIT_vs._SYSCALL/SYSRET
Newsgroups: comp.arch
Date: Wed, 20 Jan 2010 10:21:07 -0500
nmm1 writes:
Er, no. How do you stop two threads delivering the same timestamp if they execute a 'call' at the same time without having a single time server? Ensuring global uniqueness is the problem.

one the requirements was to correctly order dbms transaction log records after a failure (for recovery). a standard dbms speed-up is to allow transaction to be considered committed after the corresponding log record has been written to disk ... but the altered record in buffer memory may not be pushed out to dbms location (lazy writes to DBMS disk location).

recovery (after failure) requires using the log to sequentially "rerun" the transactions ... eventually getting the dbms image on disk to consistent state.

a cluster dbms implementation use to force record to disk before allowing it to migrate into DBMS buffer on a different processor. to speed things up, it would be possible to allow modified record to be transmitted (over high-speed link) between dbms buffers (in different processors in cluster). the problem then is that there could be multiple committed transaction changes ... recorded in different dbms logs ... but not reflected in the DBMS record.

as part of supporting direct buffer-to-buffer copies (w/o having to force out to disk) ... a mechanism was needed (for recovery) to merge transaction logs from different systems so that they have the original global temporal ordering. The requirement isn't actually to have exact time value for each transaction ... but to have multiple logs to be merged so that entries occured in the original sequence. unique accurate time works ... but so would nearly any unique monotonically increasing number (say like a transaction version number ... which could be supported as part of the operation of dbms cluster distributed lock manager ... which also piggy-backs buffer-to-buffer record copies as part of lock traffic).

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

How long for IBM System/360 architecture and its descendants?

Refed: **, - **, - **, - **
From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: How long for IBM System/360 architecture and its descendants?
Newsgroups: alt.folklore.computers
Date: Wed, 20 Jan 2010 12:21:24 -0500
Michael Wojcik <mwojcik@newsguy.com> writes:
Certainly, as Lynn has pointed out (more than once), there was some backlash against the excessive complexity of Future Systems, at least in the development of the IBM RISC processors. But in the early years of outright RISC vs. CISC competition, RISC architectures were developed on the theory that simple instruction sets with load/store architectures could be implemented with fewer, shorter cycles and deep pipelines, improving overall performance.

re:
http://www.garlic.com/~lynn/2010b.html#22 How long for IBM System/360 architecture and its descendants?
http://www.garlic.com/~lynn/2010b.html#58 How long for IBM System/360 architecture and its descendants

the instruction set simplification with 801/risc in the 70s didn't bother me so much (modulo the lack of compare&swap) ... it was the lack of cache consistency for implementating multiprocessor configuration and small number (16) of "segment" objects in the 32bit address space.

as previous note I tried to work out mechanism for packing "small shared segments" into the 801 scheme.

as long as it was pl.8 and cp.r ... the lack of hardware protection was presumably fine ... but that went out the window attempting to adapt to being unix workstation in the 80s (as well as the not supporting large number of shared memory objects).

as mentioned in some of the old 801 email ... after various (801/risc) Iliad chip efforts floundered ... some number of engineers left to work on risc at other vendors (early 80s).
http://www.garlic.com/~lynn/lhwemail.html#801

the problem with lack of compare&swap showed up with rs/6000 tho ... even tho there wasn't any multiprocessor support.

charlie had originally invented compare&swap (CAS are his initials) when working on fine-grain locking for cp67 multiprocessor. The initial attempts to get it included in 370 was rebuffed with a challenge that it needed a non-smp justification. thus was born the application multithreaded examples.
http://www.garlic.com/~lynn/subtopic.html#smp

by the time rios (rs/6000) ships ... many of the dbms implementations had adapted to compare&swap use (available on multiple different machines) ... lot more efficient than having to implement much of dbms thread serialization via kernel calls. porting various dbms to rios w/o compare&swap (even w/o multiprocessor), put rios at thruput disadvantage.

in non-multiprocessor environment ... primary semantics is being atomic and non-interruptable (not actually having to worry about serializing concurrent storage accesses). the rios (rs/6000) aix solution was special fastpath system call ... implemented in the system call interrupt routine and immediately returning (the advantage was that the system call interrupt switched to disabled for interrupts ... primarily i/o ... achieving the fundamental requirement for "atomic" compare&swap).

somerset was then started with motorola, apple, at al. the executive we reported to when doing ha/cmp
http://www.garlic.com/~lynn/subtopic.html#hacmp

moved over to head up somerset. in some sense, somerset could be considered melding rios and motorola's risc 88k for power/pc (and fixing other 801/risc trade-offs that weren't really applicable to unix & C-language environment ... or in apple's case, its unix-like mach).

lots of past posts mentioning 801, risc, iliad, romp, rios, somerset, rs/6000, power/pc, etc.
http://www.garlic.com/~lynn/subtopic.html#801

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Happy DEC-10 Day

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Happy DEC-10 Day
Newsgroups: alt.sys.pdp10, alt.folklore.computers
Date: Wed, 20 Jan 2010 13:56:46 -0500
Michael Wojcik <mwojcik@newsguy.com> writes:
Yes. Also TCP adaptive pacing requires a long-running session to reach optimal conditions. Before persistent sessions were introduced, HTTP suffered badly from TCP slow-start.

re:
http://www.garlic.com/~lynn/2010b.html#62 Happy DEC-10 Day

i think IETF meeting in aug '88(?) with presentation on slow-start ... was also the acm sigcomm meeting ... with paper on why windowing algorithms won't reach stable state in large bursty internet.

a problem in large bursty internet was to avoid large back-to-back packets at intermediate nodes overloading buffers ... and in large bursty internet ... with windowing-based algorithm ... ACKs had a tendency to bunch up on the return path. Burst of ACKs arriving all at the same time ... resulting in opening up the window and doing multiple back-to-back packet transmissions ... resulting in intermediate node congestion and overrun. the result was that things could get into pathelogical oscilation with slow-start building up size of the window ... and then having to drop back.

In that time frame ... i was doing rate-based pacing ... one of the other presentations at that IETF meeting ... was on gigabit cross-country internet ... and the amount of data in bandwidth*latency product. I had nearly the identical bandwidth*latency product on some slower speed satellite links starting a few years earlier ... and was rate-based pacing ...
http://www.garlic.com/~lynn/subnetwork.html#hsdt

also did a paper on rate-base pacing for the XTP technical advisory board.
http://www.garlic.com/~lynn/subnetwork.html#xtphsp

the other paper in the same ascm sigcomm proceedings was on tests with typical sized shielded twisted-pair ethernet networks ... showing 85% effective media thruput when all stations were in tight, low-level driver loop, constantly transmitting minimum sized packets.

we were tacking barbs from SAA and token-ring groups from our customer executive presentations on 3-tier (& ethernet).
http://www.garlic.com/~lynn/subnetwork.html#3tier

the token-ring camp was making pitches about ethernet effective thruput was less than mbit ... but I conjectured that was based on using simulation with very early 3mbit ethernet before listen-before-transmit. In any case, almaden research center had wired with CAT5 anticipating predominate 16mbit t/r deployment ... but found that 10mibt ethernet actually had both higher effective aggregate thruput as well as lower latency (over the same wires).

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Happy DEC-10 Day

Refed: **, - **, - **, - **
From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Happy DEC-10 Day
Newsgroups: alt.sys.pdp10, alt.folklore.computers
Date: Wed, 20 Jan 2010 14:36:12 -0500
Michael Wojcik <mwojcik@newsguy.com> writes:
I'm very dubious. The X.509 certificate exchange in SSL/TLS provides a lot more than the server's public key (and there would still have to be a session key exchange, etc). PKIs are not simply interchangeable like that.

Even if we assume the DNS response is trustworthy (and I'm dubious about DNSSEC's ability to achieve that), the pair {server name, public key} doesn't tell the client anything about the provenance of that key, or other identity or authorization claims being made by the server. For all the many faults of X.509, at least it's a hierarchy. Just because some DNS server tells me that foo.com has public key X doesn't mean I have any reason at all to trust a session I open using that key.


re:
http://www.garlic.com/~lynn/2010b.html#62 Happy DEC-10 Day

as part of doing this stuff with small client/server startup that wanted to do payment transactions and wanted to use this technology they had invented called "SSL" ... we had to do walk-thru/audits of various of these new organizations calling themselves Certification Authorities and issusing these things called digital certificates
http://www.garlic.com/~lynn/subpubkey.html#sslcert

SSL domain name certificate authorities have an issue with the certification process ... they aren't typically the authoritative agency for the information being certified (the information carried in the digital certificates). The problem is that the Domain Name Infrastructure is the authoritative agency for domain name ownership ... and there are some number of vulnerabilities with domain name take-over ... and then applying for valid digital certificate ... and it being granted.

Part of fall-out from DNSSEC ... for the domain name certification authority industry ... is requesting that a public key is registered as part of domain name registeration ... then future communication is digitally signed (and the DNS infrastructure can verify the digital signature with the onfile public key ... as countermeasure to domain name take-over) ... note ... there is no digital certificates involved ... misc. past posts on certificate-less public key
http://www.garlic.com/~lynn/subpubkey.html#certless

also, currently, the certification authority industry has to require a lot of identification information from domain name digital certificate applicants. They then do an error-prone, expensive, and time-consuming matching process between the supplied information and the information on file with the domain name infrastructure (as to the true owner of the domain). With onfile public keys ... they could convert to just requiring domain name digital certificate applicants to digitally sign the appication. Then the certification authority can do a real-time retrieval of the onflie public key from the domain name infrastructure ... and do a much more reliable, efficient, and less expensive signature verficiation.

there are several catch22s for the certification authority industry. first, domain name digital certificates were, in part, justified on various perceived integrity issues with the domain name infrastructure. Improving the integrity of the domain name infrastructure (such that the certification authority industry can better trust the information as part of their certification), reduces the originally justification for the digital certificates. Also, if the certification authority industry starts demonstrating that they can trust & rely on the onfile public keys ... then it is possible that others might also decide that they could rely on the DNS infrastructure, onfile public keys (further eliminating the justification for domain name digital certificates)
http://www.garlic.com/~lynn/subpubkey.html#catch22

digital certificates are analog from the letters of credit/introduction from the sailing ship days ... when the relying party had no other mechanism for information in first time dealing with perfect stranger. the original scenario for digital certificates were the offline electronic email from the early 80s ... when there would be a phone call to electronic post-office, email exchanged and then phone hung up. then a person processing the email might be faced with first time communication with complete stranger and had no other recourse to information aboth the entitty they were dealing with.

the problem as the internet became more & more pervasive and normal state of affairs was online and connected ... the original justifications for digital certificates were less & less frequently true ... and they became redundant and superfluous.

A case in point were some of the digital certificate based payment protocol specifications from the early/mid 90s. The consumer would register their public key with their financial institution and be issued a relying-party-only public key (after storing the consumer's public key in their account record). Then the consumer was expected to digitally sign every payment transaction and append their digital certificate for routing back to their financial institution. Their financial institution then retrieved the corresponding account record for executing transaction ... and would be able to verify the digital signature with the onfile public key. Appending the digital certificate represented an 100-fold payload bloat for typical payment transaction and any processing of the digital certificate represented a 100-fold processing bloat for payment transaction. That was separate from being able to trivially demonstrate that appending the digital certificate was redundant and superfluous. misc. past posts about relying-party-only digital certificates
http://www.garlic.com/~lynn/subpubkey.html#rpo

Somewhat as part of having done this stuff now called "electronic commerce", in the mid-90s, we were asked to participate in the x9a10 financial standard working group which had been given the requirement to preserve the integrity of the financial infrastrucutre for all retail payments (debit, credit, stored-value, point-of-sale, internet, face-to-face, unattended, high-value, low-value, transit turnstile ... aka ALL). The result was the x9.59 financial standard
http://www.garlic.com/~lynn/x959.html#x959

One of the things that x9.59 did was slightly tweak the paradigm and eliminated the requirement for "hiding" the transaction details (and account number) ... being able to use a digital signature ... w/o requiring an appended digital certificate.

Now the major use of "SSL" in the world today is for encryption related to hiding transaction details and account numbers ... however, with x9.59 there is no longer any requirement to hide that information ... and therefor also eliminates the major use of "SSL" in the world today.

We were tangentially involved in the cal. data breach notification legislation ... having been brought in to help word-smith the cal. electronic signature legislation ... some past posts
http://www.garlic.com/~lynn/subpubkey.html#signature

and several of the participants were also involved in privacy issues ... and had done detailed consumer studies. they found the number one issue was identity theft ... primarily account fraud identity theft resulting in fraudulent financial transactions ... and a major source of the information for those fraudulent financial transactions was coming from data breaches. Since there seemed to be little or nothing being done about data breaches ... they conjectured that publicity from mandatory data breach notification would motivate corrective action.

again, the major breaches that make it into the news involve leaking transaction details and financial account numbers. now x9.59 standard did nothing about preventing such breaches ... but it eliminated the requirement to have to hide &/or prevent account information from being divulged as part of preventing fraudulent financial transactions. x9.59 standard didn't do anything about preventing such breaches ... instead it eliminated the major common threats or exploits that might occur as result of the information leaking out.

Now there were some digital certificate oriented financial standards effort going on in parallel with x9.59. one of them recognized the enormous payload-bloat for financial transactions that comes with appending digital certificates ... so they had a standards effort to work on "compressed" digital certificates. However, using their techniques for producing "compressed" digital certificates ... I trivially showed that it was possible to compress a digital certificate to zero bytes. Then instead of x9.59 being a certificate-less protocol, it could be a digital certificate protocol, mandating that every x9.59 transaction had to include a zero-byte appended digital certificate.

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Happy DEC-10 Day

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Happy DEC-10 Day
Newsgroups: alt.sys.pdp10, alt.folklore.computers
Date: Wed, 20 Jan 2010 14:52:51 -0500
Anne & Lynn Wheeler <lynn@garlic.com> writes:
from the sailing ship days ... when the relying party had no other mechanism for information in first time dealing with perfect stranger. the original scenario for digital certificates were the offline electronic email from the early 80s ... when there would be a phone call to electronic post-office, email exchanged and then phone hung up. then a person processing the email might be faced with first time communication with complete stranger and had no other recourse to information aboth the entitty they were dealing with.

the problem as the internet became more & more pervasive and normal state of affairs was online and connected ... the original justifications for digital certificates were less & less frequently true ... and they became redundant and superfluous.


re:
http://www.garlic.com/~lynn/2010b.html#62 Happy DEC-10 Day
http://www.garlic.com/~lynn/2010b.html#69 Happy DEC-10 Day

another redundant and superfluous scenario for SSL digital certificates was as part of the original deployment of "electronic commerce" ... involving the "payment gateway"
http://www.garlic.com/~lynn/subnetwork.html#gateway

and doing SSL between the webservers and the gateway (which exchanged payment transactions between the internet and the payment networks). first, we started off mandating "mutual" authentication (which didn't exist at the time we started). Before we were done, it was also necessary to register the payment gateway with the webserver (invalidating digital certificate assumption about webserver doing first time communication with strange payment gateway) and register webservers with payment gateway (invalidating digital certificate assumption about payment gateway doing first time communication with strange webserver).

by the time everything was done and operational ... it was trivially obvious that digital certificates were redundant and superfluous ... but were installed anyway ... as as side-effect of the "SSL" public key library being used.

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Happy DEC-10 Day

Refed: **, - **, - **, - **, - **, - **, - **, - **, - **
From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Happy DEC-10 Day
Newsgroups: alt.sys.pdp10, alt.folklore.computers
Date: Wed, 20 Jan 2010 15:22:49 -0500
Michael Wojcik <mwojcik@newsguy.com> writes:
The Kingston group that took over your cluster stuff and repurposed it for number-crunching - was that the same group that had done the Scientific Visualization System / Data Explorer work? I worked on the RS/6000 version of that software (Data Explorer/6000) in my last couple of years in Cambridge.

re:
http://www.garlic.com/~lynn/2010b.html#62 Happy DEC-10 Day

one of my projects was HSDT (high-speed data transport)
http://www.garlic.com/~lynn/subnetwork.html#hsdt

it involved high-speed links ... both terrestrial and satellite and in some places HYPERChannel adapters.

there was an engineering & scientific group in kingston ... and in the early part of 80s ... I had a T1 link between san jose and the kingston engineering & scientific group over the SBS T3 IBM satellite network. There were a dozen(?) or so c-band T3 tdma IBM-dedicated earth stations that SBS had at various plant-sites ... and I had tail-circuit in san jose to the San jose earth station ... and then tail circuit from the Kingston earth station to the kingston engineering and scientific group. That E&S organization in Kinston at one point had 3090 with vector processing and numerous Floating Point Systems boxes ... and were doing things like molecular modeling. I think the scientific visualization was being done out of the E&S organization.

Then we got our own dedicated HSDT TDMA earthstations and our own dedicated transponder ... a HSDT TDMA earth station went into Yorktown on the east coast ... and the HSDT link to the Kingston E&S group switched from being circuit to the Kingston IBM earth station ... to the HSDT earth station in yorktown.

I didn't pay a lot of attention to the organization in Kingston ... but at some point there was a project to design an IBM "supercomputer" sponsored by a senior corporate executive ... it wasn't clear the lines between the (newer) supercomputer project and the kingston engineering&scientifc. The project in kingston supposedly designing a supercomputer was also providing a lot of funding to Steve Chen ... a couple recent threads mentioning Chen:
http://www.garlic.com/~lynn/2009o.html#29 Justice Department probing allegations of abuse by IBM in mainframe computer market
http://www.garlic.com/~lynn/2009s.html#5 While watching Biography about Bill Gates on CNBC last Night
http://www.garlic.com/~lynn/2009s.html#42 Larrabee delayed: anyone know what's happening?
http://www.garlic.com/~lynn/2009s.html#59 Problem with XP scheduler?

then in oct91, the senior executive sponsoring the supercomputer effort retired ... and there appeared to be serious audits of some number of projects. My impression was that was when they started looking for technology to transfer to Kingston. The Kingston organization announced a world-wide internal technology conference for mid-jan 92. We advised some of the engineers not to attend because there was possible consequences if Kingston's attention was attracted. Then things happened very, very quickly.

mostly unrelated old email regarding cluster-scaleup
http://www.garlic.com/~lynn/lhwemail.html#medusa

I was just as glad to do number crunching as dbms ... i just didn't want to do exclusively one thing ... this was about cluster scaleup for dbms
http://www.garlic.com/~lynn/95.html#13
post from today in comp.arch on some cluster scaleup issues
http://www.garlic.com/~lynn/2010b.html#66 SYSENTER/SYSEXIT_vs._SYSCALL/SYSRET

but as can be seen from this old email from end of jan92 (possibly a few hrs before the hammer fell)
http://www.garlic.com/~lynn/2006x.html#email920129

I was also involved with national labs on cluster scaleup that didn't involve dbms work

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Happy DEC-10 Day

Refed: **, - **, - **, - **, - **, - **, - **, - **
From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Happy DEC-10 Day
Newsgroups: alt.sys.pdp10, alt.folklore.computers
Date: Wed, 20 Jan 2010 17:14:58 -0500
Anne & Lynn Wheeler <lynn@garlic.com> writes:
group. That E&S organization in Kinston at one point had 3090 with vector processing and numerous Floating Point Systems boxes ... and were doing things like molecular modeling. I think the scientific visualization was being done out of the E&S organization.

re:
http://www.garlic.com/~lynn/2010b.html#71 Happy DEC-10 Day

old post mentioning jun86, Kingston E&S had 20 FPS "attached processor" boxes (aggregate peak 1.6gflop):
http://www.garlic.com/~lynn/2000c.html#61 TF-1

some drift, the following post:
http://www.garlic.com/~lynn/2000c.html#62 ANSI X9.62 and X9.63

for a little x-over with
http://www.garlic.com/~lynn/2010b.html#69 Happy DEC-10 Day

was somebody asking for copies of x9.62 & x9.63 ansi standards ... one of which was the work on compressed digital certificates (and the other on elliptical curve cryptography).

and even more drift (a couple posts down in the same archive)
http://www.garlic.com/~lynn/2000c.html#65 Does the word "mainframe" still have meaning?

discusses doing HYPERChannel work for the IMS groups in STL and Boulder (in 80/81). The IT guy I was working with in Boulder then transfers to Kingston E&S group ... and I work with him on the HSDT link into the E&S operation.
http://www.garlic.com/~lynn/subnetwork.html#hsdt

The HYPERChannel work for the IMS groups in STL and Boulder was unrelated to when Jim left for Tandem, his palming off DBMS consulting to the IMS group ... old email refs:
http://www.garlic.com/~lynn/2007.html#email801006
http://www.garlic.com/~lynn/2007.html#email801016

misc. other posts mentioning Floating Point System boxes:
http://www.garlic.com/~lynn/2000c.html#5 TF-1
http://www.garlic.com/~lynn/2001b.html#56 Why SMP at all anymore?
http://www.garlic.com/~lynn/2001m.html#25 ESCON Data Transfer Rate
http://www.garlic.com/~lynn/2002.html#0 index searching
http://www.garlic.com/~lynn/2002e.html#31 Hardest Mistake in Comp Arch to Fix
http://www.garlic.com/~lynn/2002j.html#30 Weird
http://www.garlic.com/~lynn/2003b.html#29 360/370 disk drives
http://www.garlic.com/~lynn/2003d.html#35 Why only 24 bits on S/360?
http://www.garlic.com/~lynn/2003m.html#20 360 Microde Floating Point Fix
http://www.garlic.com/~lynn/2006m.html#4 The Power of the NORC

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Happy DEC-10 Day

Refed: **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **
From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Happy DEC-10 Day
Newsgroups: alt.sys.pdp10, alt.folklore.computers
Date: Wed, 20 Jan 2010 17:50:28 -0500
johnf@panix.com (John Francis) writes:
The company I work for sells airline schedule lookup tools that can be incorporated into a web page. We take in a few GB of data on a daily basis, digest it down to a smaller format (anywhere from 1MB to 10MB), and have a wicked fast lookup engine running on our servers.

recent post mentioning being asked to consult with large airline res system ... and coming back with rewritten "routes" ... involved taking everything from the full machine readable industry OAG (at the time over 4000 airports world-wide & over 400k flt segments ... or 600k flts ... depending on how things were counted ... some equipment flew with multiple flt designations) .... digested down to much smaller format ... and did wicked fast loopup on nearly any machine with at least 32mbytes of real storage (would operate as interactive PC application or as client/server operation).
http://www.garlic.com/~lynn/2010b.html#13 Larrabee delayed: anyone know what's happening?

At the time, "routes" represented 25% of the processing load on their res. system. As mentioned, the initial pass ran 20 times faster than their mainframe implementation ... and for version running on rs/6000 320 ... some careful processing organization for 6000 cache sensitivity ... got another five times improvement (overall 100 times improvement). Then redid the sequence ... so that several human interactions were collapsed into single interaction. That single interaction then was only about ten times faster than any of the individual original interactions.

As mentioned it could run on just about any PC or workstation with 32mbytes of real storage ... and ran either as interactive application or configured to run client/server.

misc. past posts mentioning redoing routes (took two months elapsed time for the redesign/rewrite):
http://www.garlic.com/~lynn/99.html#136a checks (was S/390 on PowerPC?)
http://www.garlic.com/~lynn/2000f.html#20 Competitors to SABRE?
http://www.garlic.com/~lynn/2001d.html#69 Block oriented I/O over IP
http://www.garlic.com/~lynn/2002g.html#2 Computers in Science Fiction
http://www.garlic.com/~lynn/2002j.html#83 Summary: Robots of Doom
http://www.garlic.com/~lynn/2004b.html#6 Mainframe not a good architecture for interactive workloads
http://www.garlic.com/~lynn/2004o.html#23 Demo: Things in Hierarchies (w/o RM/SQL)
http://www.garlic.com/~lynn/2004q.html#85 The TransRelational Model: Performance Concerns
http://www.garlic.com/~lynn/2006o.html#18 RAMAC 305(?)
http://www.garlic.com/~lynn/2007g.html#22 Bidirectional Binary Self-Joins
http://www.garlic.com/~lynn/2007g.html#41 US Airways badmouths legacy system
http://www.garlic.com/~lynn/2007h.html#41 Fast and Safe C Strings: User friendly C macros to Declare and use C Strings
http://www.garlic.com/~lynn/2008i.html#19 American Airlines
http://www.garlic.com/~lynn/2008j.html#32 CLIs and GUIs
http://www.garlic.com/~lynn/2008p.html#41 Automation is still not accepted to streamline the business processes... why organizations are not accepting newer technologies?

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Happy DEC-10 Day

Refed: **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **
From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Happy DEC-10 Day
Newsgroups: alt.sys.pdp10, alt.folklore.computers
Date: Wed, 20 Jan 2010 19:00:46 -0500
johnf@panix.com (John Francis) writes:
We do the front-end stuff (the stuff that Orbitz/Expedia/... do); find which of all those flight segments (almost 2M of them nowadays) make sense for a particular itinerary (including multiple flight segments, and taking into account Minimum Connect Times and Traffic Restriction Codes). You say you want to get from Strasbourg to Ann Arbor, say, and we take it from there.

re:
http://www.garlic.com/~lynn/2010b.html#73 Happy DEC-10 Day

I took it back to demo ... and they had some things that they considered hard to do ... they asked for some unknown origin airport in kansas and some equally unknown destination airport in georgia; instantaneously found route that involved five connections and move than 24hrs elapsed time (hint: kansas and georgia weren't in the same country) ... also most minimum connect times. This was something at the time, that they couldn't do. It was actually faster than for doing SFO to Kennedy ... because there are much larger number of possibilities.

I had minimum connect times for most airports between different gates in the same airport ... what i didn't have in that first version was minimum connect times for New York or Washington ... both are "generic" names for multiple airports ... where arrival at one gate may then involve 90min bus ride to connection out of a different gate (plus redoing security).

also got a separate file that gave latitude/longitude for every airport in the world ... and so could draw simplistic (straight line) for flt segments between airports.

part of the difference was that the traditional "routes" had evolved from the 60s as a lookup of database with predetermined ways of getting from A-to-B.

I had recently done a port of 60k statement vs/pascal program to other vendors (mostly workstation) platforms (part of IBM moving off of large set of internal proprietary vlsi design tools to industry tools ... one of the approaches was giving the internal tools to industry tool companies) ... which did automated physical layout. While I did the airline res in C rather than pascal .. there were some similarities ... although the airline res was actually easier.

the layout porting was interesting ... because some of these other (workstation) vendor pascals ... appeared to never had been used for other than educational institution student assignments; which in one case was complicated by the vendor had outsourced their pascal support to organization 12 time zones away (so even tho I could easily drop in on the workstation vendor ... there was still at least 24hr turn-around on pascal issues).

past posts mentioning porting the layout
http://www.garlic.com/~lynn/99.html#213 Why is Pascal no longer a leading development Language?
http://www.garlic.com/~lynn/2001b.html#30 perceived forced conversion from cp/m to ms-dos in late 80's
http://www.garlic.com/~lynn/2001e.html#76 Stoopidest Hardware Repair Call?
http://www.garlic.com/~lynn/2002q.html#19 Beyond 8+3
http://www.garlic.com/~lynn/2004f.html#42 Infiniband - practicalities for small clusters
http://www.garlic.com/~lynn/2004k.html#34 August 23, 1957
http://www.garlic.com/~lynn/2004q.html#35 [Lit.] Buffer overruns
http://www.garlic.com/~lynn/2005b.html#14 something like a CTC on a PC
http://www.garlic.com/~lynn/2005o.html#11 ISA-independent programming language
http://www.garlic.com/~lynn/2007j.html#16 Newbie question on table design
http://www.garlic.com/~lynn/2007o.html#61 (Newbie question)How does the modern high-end processor been designed?
http://www.garlic.com/~lynn/2007u.html#16 Fazing out x86
http://www.garlic.com/~lynn/2008j.html#77 CLIs and GUIs

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Windows plagued by 17-year-old privilege escalation bug

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Windows plagued by 17-year-old privilege escalation bug
Newsgroups: alt.folklore.computers
Date: Wed, 20 Jan 2010 19:14:37 -0500
Windows plagued by 17-year-old privilege escalation bug; All 32-bit versions vulnerable
http://www.theregister.co.uk/2010/01/19/microsoft_escalation_bug/

from above:
The vulnerability resides in a feature known as the Virtual DOS Machine, which Microsoft introduced in 1993 with Windows NT, according to this writeup penned by Tavis Ormandy of Google. Using code written for the VDM, an unprivileged user can inject code of his choosing directly into the system's kernel, making it possible to make changes to highly sensitive parts of the operating system.

... snip ...

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Happy DEC-10 Day

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Happy DEC-10 Day
Newsgroups: alt.sys.pdp10, alt.folklore.computers
Date: Wed, 20 Jan 2010 19:43:52 -0500
Anne & Lynn Wheeler <lynn@garlic.com> writes:
the layout porting was interesting ... because some of these other (workstation) vendor pascals ... appeared to never had been used for other than educational institution student assignments; which in one case was complicated by the vendor had outsourced their pascal support to organization 12 time zones away (so even tho I could easily drop in on the workstation vendor ... there was still at least 24hr turn-around on pascal issues).

re:
http://www.garlic.com/~lynn/2010b.html#74 Happy DEC-10 Day

old post from 2006 ... somebody had recently been near the location above and brought me back a souvenir
http://www.garlic.com/~lynn/2006r.html#48 cold war again

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

search engine history, was Happy DEC-10 Day

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: search engine history, was Happy DEC-10 Day
Newsgroups: alt.sys.pdp10, alt.folklore.computers
Date: Wed, 20 Jan 2010 20:52:16 -0500
Peter Flass <Peter_Flass@Yahoo.com> writes:
Plus all the RS-6000's, i-series boxes (AS/400), and ther many PPC chips in IBM's mainframes and controllers.

re:
http://www.garlic.com/~lynn/2010.html#42 Happy DEC-10 Day
http://www.garlic.com/~lynn/2010.html#43 Happy DEC-10 Day
http://www.garlic.com/~lynn/2010b.html#22 How long for IBM System/360 architecture and its descendants?
http://www.garlic.com/~lynn/2010b.html#58 How long for IBM System/360 architecture and its descendants?
http://www.garlic.com/~lynn/2010b.html#67 How long for IBM System/360 architecture and its descendants?

ppc processors
http://en.wikipedia.org/wiki/PowerPC

above also mentions F-35

and ppc cores in cell processors
http://en.wikipedia.org/wiki/Cell_microprocessor
in game consoles
http://en.wikipedia.org/wiki/Playstation_3
and
http://en.wikipedia.org/wiki/Xbox_360

also
http://en.wikipedia.org/wiki/Gekko_(microprocessor)
in
http://en.wikipedia.org/wiki/GameCube

also
http://en.wikipedia.org/wiki/Broadway_(microprocessor)
in
http://en.wikipedia.org/wiki/Wii

misc. past posts mentioning 801, risc, romp, rios, somerset, etc
http://www.garlic.com/~lynn/subtopic.html#801

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Happy DEC-10 Day

Refed: **, - **, - **, - **, - **, - **
From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Happy DEC-10 Day
Newsgroups: alt.sys.pdp10, alt.folklore.computers
Date: Thu, 21 Jan 2010 09:29:12 -0500
Anne & Lynn Wheeler <lynn@garlic.com> writes:
At the time, "routes" represented 25% of the processing load on their res. system. As mentioned, the initial pass ran 20 times faster than their mainframe implementation ... and for version running on rs/6000 320 ... some careful processing organization for 6000 cache sensitivity ... got another five times improvement (overall 100 times improvement). Then redid the sequence ... so that several human interactions were collapsed into single interaction. That single interaction then was only about ten times faster than any of the individual original interactions.

As mentioned it could run on just about any PC or workstation with 32mbytes of real storage ... and ran either as interactive application or configured to run client/server.


re:
http://www.garlic.com/~lynn/2010b.html#13 Larrabee delayed: anyone know what's happening?
http://www.garlic.com/~lynn/2010b.html#73 Happy DEC-10 Day
http://www.garlic.com/~lynn/2010b.html#74 Happy DEC-10 Day

small configuration is relative ... the above was done over 15yrs ago

old dhrystone table:


                                                      CPU   MIPS   MIPS
    System                      OS          CPU     (MHz)  V1.1   V2.1  REF
003 Dell Dimension Pro150  NT 4.0 srvr  Pentium Pro 200.0 ------  446.9 103
177 IBM RS/6000 Model 320  ------------ Power RISC   20.0   29.5   25.8   5

however this reference ..
http://en.wikipedia.org/wiki/Instructions_per_second

has 200mhs Intel Pentium Pro at 541mips

cache size can also make a difference ... I have an application that gets a little bit better thruput on 2mbyte cache 1.7ghz machine as on a 512kbyte cache 3.4ghz machine.

in any case, current processors are 1000 times or more faster than the 320 ... also from above:


2.93 ghz intel core 2 extreme x6800     27,079mips
3.2  ghz intel core 2 extreme qx9770    59,455mips

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Happy DEC-10 Day

Refed: **, - **, - **, - **, - **, - **, - **, - **
From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Happy DEC-10 Day
Newsgroups: alt.sys.pdp10, alt.folklore.computers
Date: Thu, 21 Jan 2010 09:54:57 -0500
re:
http://www.garlic.com/~lynn/2010b.html#78 Happy DEC-10 Day

for the fun of it ... the old dhrystone table:
http://performance.netlib.org/performance/html/dhrystone.data.col0.html

other pieces of the table ...


                                                     CPU   MIPS   MIPS
    System                      OS          CPU     (MHz)  V1.1   V2.1  REF
### ---------------------- ------------ ----------- ----- ------ ------ ---

001 AlphaServer 8400 5/300 UNIX V4.0    DEC 21164   300.0  550.1  464.8  99
002 AlphaServer 8400 5/300 UNIX V4.0    DEC 21164   300.0  523.5  457.2  97
003 Dell Dimension Pro150  NT 4.0 srvr  Pentium Pro 200.0 ------  446.9 103
004 Dell Dimension Pro150  NT 4.0 srvr  Pentium Pro 200.0 ------  419.5 101
005 DEC Alpha 600 5/266    OSF/1 V3.2c  21164-EB5   266.0 ------  366.8  79
006 DEC Server 2100 5/250  UNIX V3.2b   DEC 21064   250.0 ------  360.4  70
007 Dell XPS Pro 200n      NT 3.51      Pentium Pro 200.0  372.8  312.4 110
008 DEC 3000/900  AXP      OSF/1 V3.0   DEC 21064   275.0 ------  291.9  63
009 DEC Alpha 600 5/266    OSF/1 V3.2c  21164-EB5   266.0 ------  290.0  79

... snip ...

and xscale processor used in treo
http://en.wikipedia.org/wiki/XScale

from above: 800 MIPS for 624MHz PXA270 vs. 1000 MIPS for 1.25 GHz Monahans

say 30 times or more faster than rs6000/320.

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Happy DEC-10 Day

Refed: **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **
From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Happy DEC-10 Day
Newsgroups: alt.sys.pdp10, alt.folklore.computers
Date: Thu, 21 Jan 2010 13:35:55 -0500
johnf@panix.com (John Francis) writes:
Note that memory speed doesn't scale quite like that on that platform. The memory technology used on the treo is disproportionately slow.

re:
http://www.garlic.com/~lynn/2010b.html#73 Happy DEC-10 Day
http://www.garlic.com/~lynn/2010b.html#74 Happy DEC-10 Day
http://www.garlic.com/~lynn/2010b.html#78 Happy DEC-10 Day
http://www.garlic.com/~lynn/2010b.html#79 Happy DEC-10 Day

Looking at old reference from the archives, I mention that based on rs6000/320 thruput measurements ... indicates that if moved to rs6000/580, it would trivially handle their existing transaction load running on muliple 3090s ... and their target expanded load requiring several hundred ES9000 processors could be handled on ten 580s (changing things from DBMS lookup to purely compute route finding); this involved doing more processing (as I was demonstrating in rewritten application), more flts, and significant increase in number of transactions.

for one of the demos, there was 4048 airports and 635820 flt segments.

I don't remember 580 mip rate off-hand ... but from the dhrystone data, a 580 would be somewhere between 550 and 590.
http://performance.netlib.org/performance/html/dhrystone.data.col0.html


 51 IBM RS/6000 590        AIX 3.2.5    Power2      -----  134.2  124.3  47
108 IBM RS/6000 Model 550  AIX 3.2.2    Power RISC   42.7   80.8   62.9   2
177 IBM RS/6000 Model 320  ------------ Power RISC   20.0   29.5   25.8   5

say 580 possibly in range of 100mips ... then a 1000mip TREO potentially is the equivalent of the ten 580s (theoritically full, world-wide transaction load) ... if it had faster memory.

I had a thing about about "change of equipment". The earliest I'm aware of is early morning TWA flt out of San Jose.

Printed OAG and reservation terminals would list non-stop and direct flts before listing connecting flts. The airlines came up with "change of equipment" ... a flt takes off with multiple flt numbers going to different destinations. It lands at some point, and some passengers have to get off for "change of equipment" (connection by any other name). This got more of an airlines flts listed in the (top) "direct" section (as opposed to the connecting section).

TWA used to park some planes overnight in San Jose because it was cheaper than SFO. Early morning TWA flt out of san jose left with two flt. numbers, one going to Seattle and one going to Kennedy. Passengers going to Seattle would stay on the plane when it stopped in SFO ... but passengers going to Kennedy had to get off in SFO and change to a different plane.

There appeared to be a Contential flt Honolulu to LAX with the most flt numbers (and most change of equipment) ... aka half dozen Contential flts numbers all listed as departing Honolulu at the same time and arrive LAX at the same time (and flying same model plane). Of course today ... with alliances ... not only will same equipment might have multiple flt. numbers for the same carrier ... but might also have multiple flt. numbers for multiple different carriers.

One of the other issues for airline res system was that it only had information for a limited number of connections ... for origin/destination requiring more connections ... it had to be figured manually. An excuse for some of the multiple flt number scheme was it made it easier for agents to work out some of the more complex origin/destination. My application rewrite ... would follow as many connections as necessary in order to get between two points (and eliminated that justification for flts with multiple flt numbers).

Other random trivia ... there was flt. in South America with largest number of flt. segments ... i.e. flt departs first thing in the morning ... makes more than dozen landing/take-offs during the course of the day ... before landing that night at the same airport in started the day from.

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Happy DEC-10 Day

Refed: **, - **, - **, - **, - **, - **, - **, - **, - **, - **
From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Happy DEC-10 Day
Newsgroups: alt.sys.pdp10, alt.folklore.computers
Date: Fri, 22 Jan 2010 00:16:07 -0500
Pat Farrell <pfarrell@pfarrell.com> writes:
Early this century, I worked for a while at Fannie Mae. They had a small problem with rounding of numbers. It made the front page of the Washington Post, NY Times and Wall Street Journal.

The error was $900 million dollars. The press made out that this was a big problem. It was really just rounding at the sums that Fannie Mae dealt with.

The problem was more fundamental that even improper use of floating point. They had some systems written as Excel macros. They had systems written on ten if not twenty different kinds of hardware, and too many kinds of software to count.

The corporate data processing folks were so far behind the curve that the analysts just implemented what they needed locally. Including in Excel.

Sigh.


re:
http://www.garlic.com/~lynn/2009.html#73 CROOKS and NANNIES: what would Boyd do?

... from above post ...

Also, GAO has started doing a database of executives fiddling public company financial reports (in spite of SOX). The executives get a boost in compensation based on the fiddled numbers. Later the financials may be restated ... but the compensation not forfeited. One example was in 2004 Freddie was fined $400m for $10b fiddling of financials and the CEO replaced ... but allowed to keep $60m.

GAO references:
http://www.gao.gov/products/GAO-03-138
and
http://www.gao.gov/new.items/d06678.pdf

... snip ...

there was also an earlier CBS news item that supposedly at one point Freddie had more lobbiests (many who were former congressmen) on its payroll than employess.

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Oldest Instruction Set still in daily use?

Refed: **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **
From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Oldest Instruction Set still in daily use?
Newsgroups: alt.folklore.computers
Date: Fri, 22 Jan 2010 10:28:20 -0500
jmfbahciv <jmfbahciv@aol> writes:
A perfect example why Congress is such a mess. We are getting what we vote for.

GLBA was pretty good example of "buying" votes with special provisions (as countermeasure to presidential veto)

folklore from the time of GLBA was that president was going to veto the bill ... republicans easily had the votes to pass the bill, but not override the veto. then there were provisions added to the bill that eventually got sufficient dem votes to easily override any veto ... at which point they passed the bill, sent it to president, and president signed the bill (veto on bill with such a lopsided vote would have been pointless). somewhat supported by the wiki write-up ... but presented/phrased somewhat differently (and didn't go into lots of detail about what provisions for which votes; not like some of the very state-specific stuff in health bill)
http://en.wikipedia.org/wiki/Gramm-Leach-Bliley_Act

about the time that GLBA was going on ... we had been brought in to help word smith the cal. state electronic signature legislation .. and I've mentioned several times being tangentially involved in the cal. state data breach notification legislation. some past posts
http://www.garlic.com/~lynn/subpubkey.html#signature

Several participants in electronic signature legislation were also heavily involved in privacy issues. there had been detailed, in-depth consumer surveys and the number one privacy issue was identity theft ... specifically kind resulting in fraudulent financial transactions (account fraud) ... frequently as a result of some data breach. There seemed to be little or no corrective action being done about the situation; so they appeared to believe that resulting publicity from data breach notifications would help motivate corrective action.

Besides repeal of Glass-Steagall, there was big deal that GLBA (1999 bank modernization act) was specifically going to prevent walmart (and m'soft) from becoming banks (if you already are a bank, you get to stay a bank; if you aren't already a bank, you don't get to become one) as a mechanism for protecting small community banks (they may have more to worry about from the too-big-to-fail institutions than walmart).

Cal was also preparing an "opt-in" cal. privacy legislation ... when "opt-out" was added to GLBA. In "opt-in", the consumer has to specifically authorize sharing of personal information. In GLBA, "opt-out" allows sharing unless the consumer has notified institution that they didn't want sharing ("opt-in" was viewed as being significantly more onerous to financial industry ... and people in cal. viewed the addition of "opt-out" to GLBA as federal pre-emption of their efforts; they also expressed concern about what other things congress might do in the way of "federal pre-emption").

Later (2003 or 2004), I was at national privacy conference meeting at Renaissance hotel in Washington DC. One of the sessions had panel of the FTC commissioners. In the Q&A, somebody in the audience got up and said he worked on customer callcenter software used by most of the financial industry. He claimed he knew that most of the people in call-centers answering 1-800 number for "opt-out", were not provided any mechanism for recording caller information (no record was kept of callers wanting to opt-out). He asked the FTC commissioners if they had any intention of investigating the situation.

25 People to Blame for the Financial Crisis; Phil Gramm
http://content.time.com/time/specials/packages/article/0,28804,1877351_1877350_1877330,00.html
PBS program looking at repeal of Glass-Steagall
http://www.pbs.org/wgbh/pages/frontline/shows/wallstreet/weill/demise.html
Glass-Steagall wiki page
http://en.wikipedia.org/wiki/Glass-Steagall_Act

and then commodoty futures modernization act

The Warning
http://www.pbs.org/wgbh/pages/frontline/warning/
Interview: Brooksley Born
http://www.pbs.org/wgbh/pages/frontline/warning/interviews/born.html
Gramm and the 'Enron Loophole'
http://www.nytimes.com/2008/11/17/business/17grammside.html
Phil Gramm's Enron Favor
http://www.villagevoice.com/2002-01-15/news/phil-gramm-s-enron-favor/
Greenspan Slept as Off-Books Debt Escaped Scrutiny
http://www.bloomberg.com/apps/news?pid=20601109&refer=home&sid=aYJZOB_gZi0I

a few posts from last year mentioning above:
http://www.garlic.com/~lynn/2009c.html#38 People to Blame for the Financial Crisis
http://www.garlic.com/~lynn/2009c.html#49 How to defeat new telemarketing tactic
http://www.garlic.com/~lynn/2009c.html#53 How to defeat new telemarketing tactic
http://www.garlic.com/~lynn/2009c.html#55 Who will give Citigroup the KNOCKOUT blow?
http://www.garlic.com/~lynn/2009c.html#65 is it possible that ALL banks will be nationalized?
http://www.garlic.com/~lynn/2009d.html#10 Who will Survive AIG or Derivative Counterparty Risk?
http://www.garlic.com/~lynn/2009d.html#28 I need insight on the Stock Market
http://www.garlic.com/~lynn/2009d.html#61 Quiz: Evaluate your level of Spreadsheet risk
http://www.garlic.com/~lynn/2009d.html#62 Is Wall Street World's Largest Ponzi Scheme where Madoff is Just a Poster Child?
http://www.garlic.com/~lynn/2009d.html#63 Do bonuses foster unethical conduct?
http://www.garlic.com/~lynn/2009d.html#73 Should Glass-Steagall be reinstated?
http://www.garlic.com/~lynn/2009e.html#0 What is swap in the financial market?
http://www.garlic.com/~lynn/2009e.html#8 The background reasons of Credit Crunch
http://www.garlic.com/~lynn/2009e.html#13 Should we fear and hate derivatives?
http://www.garlic.com/~lynn/2009e.html#23 Should FDIC or the Federal Reserve Bank have the authority to shut down and take over non-bank financial institutions like AIG?
http://www.garlic.com/~lynn/2009e.html#35 Architectural Diversity
http://www.garlic.com/~lynn/2009f.html#29 What is the real basis for business mess we are facing today?
http://www.garlic.com/~lynn/2009f.html#38 On whom or what would you place the blame for the sub-prime crisis?
http://www.garlic.com/~lynn/2009f.html#51 On whom or what would you place the blame for the sub-prime crisis?
http://www.garlic.com/~lynn/2009f.html#53 What every taxpayer should know about what caused the current Financial Crisis
http://www.garlic.com/~lynn/2009g.html#5 Do the current Banking Results in the US hide a grim truth?
http://www.garlic.com/~lynn/2009g.html#7 Just posted third article about toxic assets in a series on the current financial crisis
http://www.garlic.com/~lynn/2009g.html#33 Treating the Web As an Archive
http://www.garlic.com/~lynn/2009g.html#76 Undoing 2000 Commodity Futures Modernization Act
http://www.garlic.com/~lynn/2009h.html#17 REGULATOR ROLE IN THE LIGHT OF RECENT FINANCIAL SCANDALS
http://www.garlic.com/~lynn/2009i.html#54 64 Cores -- IBM is showing a prototype already
http://www.garlic.com/~lynn/2009i.html#60 In the USA "financial regulator seeks power to curb excess speculation."
http://www.garlic.com/~lynn/2009i.html#74 Administration calls for financial system overhaul
http://www.garlic.com/~lynn/2009i.html#77 Financial Regulatory Reform - elimination of loophole allowing special purpose institutions outside Bank Holding Company (BHC) oversigh
http://www.garlic.com/~lynn/2009j.html#21 The Big Takeover
http://www.garlic.com/~lynn/2009j.html#30 An Amazing Document On Madoff Said To Have Been Sent To SEC In 2005
http://www.garlic.com/~lynn/2009j.html#35 what is mortgage-backed securities?
http://www.garlic.com/~lynn/2009l.html#5 Internal fraud isn't new, but it's news
http://www.garlic.com/~lynn/2009n.html#56 Opinions on the 'Unix Haters' Handbook'
http://www.garlic.com/~lynn/2009o.html#84 Opinions on the 'Unix Haters' Handbook'
http://www.garlic.com/~lynn/2009p.html#51 Opinions on the 'Unix Haters' Handbook
http://www.garlic.com/~lynn/2009q.html#77 Now is time for banks to replace core system according to Accenture

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Happy DEC-10 Day

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Happy DEC-10 Day
Newsgroups: alt.folklore.computers
Date: Fri, 22 Jan 2010 10:48:09 -0500
Natarajan Krishnaswami <natarajan+usenet@krishnaswami.org> writes:
Floating point has desirable properties for exponential processes, such as interest, statistical calculations, and depreciation. (Things like the decimal floating point Cowlishaw (ObAFC: Rexx!) has been pushing), even more so.)

Mike had presentation at Hillgang meeting 2 yrs ago:
http://www.mail-archive.com/linux-390@vm.marist.edu/msg47531.html

copy of the presentation:

Changing the Way Computers Compute
http://www.sinenomine.net/publications/conference/hillgang/cowlishaw-200804

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Happy DEC-10 Day

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Happy DEC-10 Day
Newsgroups: alt.sys.pdp10, alt.folklore.computers
Date: Fri, 22 Jan 2010 11:20:03 -0500
Pat Farrell <pfarrell@pfarrell.com> writes:
I don't think that is literally true. But it is true that all the IT folks moved out of the fancy Washington DC office to way out near Dulles. They had an IT staff of nearly 1000. I don't know if all of them were employees, some were consultants.

The fancy DC building was very heavily PR and Legal folks.

There was a big revolving door, and the big wigs were very well connected to the political powers.


re:
http://www.garlic.com/~lynn/2010b.html#81 Happy DEC-10 Day

the story was something about a $20,000/annum default lobbyiest retainer for nearly everybody that had ever been anybody in the washington area; 1000 is measly $20m/annum, 5000 is still only $100m/annum (about the same as the CEO's compensation) ... they didn't necessarily actually have to do anything ... just be oncall if they were needed.

this is 2006 list with freddie mac at 5000 employees
http://money.cnn.com/magazines/fortune/global500/2006/snapshots/543.html

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

search engine history, was Happy DEC-10 Day

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: search engine history, was Happy DEC-10 Day
Newsgroups: alt.sys.pdp10, alt.folklore.computers
Date: Fri, 22 Jan 2010 13:59:36 -0500
Michael Wojcik <mwojcik@newsguy.com> writes:
The original RS/6000 (RIOS) used a 6-chip CPU board, which was the first commercial POWER ("Performance Optimized With Enhanced RISC", where RISC was redefined as "Reduced Instruction Set Cycles", according to Phil Hester).

I've got RIOS six chip clear plastic paper weight right now on my desk ... picture of it here
http://www.garlic.com/~lynn/lhwemail.html#oldpicts

it was also claimed that rios had a 52bit address ... this was left over from romp claiming 40bit address ... and left over from the original 801/risc philosophy.

rios/romp/et.al had 32bit instruction addresses; top four bits mapped to one of sixteen segment registers with low 28-bit segment offset.

801 had inverted tables and virtual addresses were segment-id associative. In romp, the segment register was 12bits ... allowing concurrent definition of 4096 (virtual) segments. from pre-unix 801/romp lore ... since inline application code could change segment-id in register as easily as address in general purpose register ... there was claim applications had a 12+28 bit effective address space (the 12bit segment-id plus the 28bit segment offset).

moving into unix with a more traditional 32bit addressing paradigm ... with relatively statically assigned segments ... there tended to be relatively static segment-id values to simulate a single virtual address space.

for instance, in 370/168 ... it was address-space associative. there was a 7-entry "STO-stack" ... and then each (virtual memory) table-look-aside entry had a 3-bit (virtual address space) identifier. STO is "segment table origin" address ... where there was a unique segment table per address space. There could be an arbitrary large number of virtual address spaces (as many segment tables that could be built in real memory) ... but the 168 only remembered the most recent seven.

The 801/risc architecture eliminated the hardware having to manage all such stuff ... and pushed it off on software.

In the move from ROMP to RIOS ... even tho by then it was purely UNIX platform ... somehow they managed to retain the pre-unix ROMP "addressing" description ... except that RIOS now had 24bit segment registers (which theoritically gave 24bit+28bit=52bit ... assuming the non-unix, earlier cp.r programming paradigm).

misc. past 801, risc, iliad, romp, rios, power, power/pc, somerset, etc
http://www.garlic.com/~lynn/subtopic.html#801

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Oldest Instruction Set still in daily use?

Refed: **, - **, - **, - **, - **, - **, - **
From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Oldest Instruction Set still in daily use?
Newsgroups: alt.folklore.computers
Date: Fri, 22 Jan 2010 14:13:26 -0500
Anne & Lynn Wheeler <lynn@garlic.com> writes:
Besides repeal of Glass-Steagall, there was big deal that GLBA (1999 bank modernization act) was specifically going to prevent walmart (and m'soft) from becoming banks (if you already are a bank, you get to stay a bank; if you aren't already a bank, you don't get to become one) as a mechanism for protecting small community banks (they may have more to worry about from the too-big-to-fail institutions than walmart).

re:
http://www.garlic.com/~lynn/2010b.html#82 Oldest Instruction Set still in daily use?

the issue with walmart was that walmart accounted for somewhere between 25-30% of the retail transactions in the country ... and corresponding percentage of plastic card payment transactions. the interchange fee taken by the financial infrastructure on those transactions are enormous. walmart was claiming it wanted a bank charter to become its own merchant acquiring bank ... being able to retain that portion of payment transaction interchange fee. that possibility, easily justified the amount of money that was poured into congress for GLBA ... some news report last year that the financial industry got $250,000 in benefits for every dollar they spent in contributions and lobbying
http://www.garlic.com/~lynn/2009p.html#2 Opinions on the 'Unix Haters' Handbook

Requlated portion of the too-big-to-fail US financial institutions have gotten 40-60% of their bottom line from these fees ... and walmart being able to retain the merchant interchange fees would be a big blow (to those too-big-to-fail merchant acquiring institutions).

the scenario that these institutions used with the community banks, was the notorious reputation that walmart has for efficiency and cutting fat and overhead ... and what would happen if they did it for banking. currently about 1/3rd of the population are unbanked ... mostly below the profit margin for current financial institution operations. Specter is that walmart would come in and apply its reputation for cutting fat and overhead to banking and then be able to profitably provide financial services to those unbanked. Once walmart had a lean & mean profitable financial operation with 1/3rd of the country as their customer base ... it would put all the other financial institutions at a competitive disadvantage (potentially forcing them to also transition to operation with enormously reduced fat and overhead).

misc. past posts mentioning interchange fees
http://www.garlic.com/~lynn/2005u.html#16 AMD to leave x86 behind?
http://www.garlic.com/~lynn/2006k.html#23 Value of an old IBM PS/2 CL57 SX Laptop
http://www.garlic.com/~lynn/2007.html#27 Securing financial transactions a high priority for 2007
http://www.garlic.com/~lynn/2007c.html#38 Securing financial transactions a high priority for 2007
http://www.garlic.com/~lynn/2007h.html#56 T.J. Maxx data theft worse than first reported
http://www.garlic.com/~lynn/2007i.html#17 John W. Backus, 82, Fortran developer, dies
http://www.garlic.com/~lynn/2007i.html#47 John W. Backus, 82, Fortran developer, dies
http://www.garlic.com/~lynn/2007i.html#59 John W. Backus, 82, Fortran developer, dies
http://www.garlic.com/~lynn/2007i.html#72 Free Checking
http://www.garlic.com/~lynn/2007l.html#35 My Dream PC -- Chip-Based
http://www.garlic.com/~lynn/2007n.html#68 Poll: oldest computer thing you still use
http://www.garlic.com/~lynn/2007r.html#31 Is the media letting banks off the hook on payment card security
http://www.garlic.com/~lynn/2007r.html#40 Is the media letting banks off the hook on payment card security
http://www.garlic.com/~lynn/2007s.html#64 Is the media letting banks off the hook on payment card security
http://www.garlic.com/~lynn/2007u.html#0 folklore indeed
http://www.garlic.com/~lynn/2007v.html#62 folklore indeed
http://www.garlic.com/~lynn/2008c.html#7 Toyota Sales for 2007 May Surpass GM
http://www.garlic.com/~lynn/2008c.html#90 Toyota Sales for 2007 May Surpass GM
http://www.garlic.com/~lynn/2008d.html#3 Govt demands password to personal computer
http://www.garlic.com/~lynn/2008j.html#58 dollar coins
http://www.garlic.com/~lynn/2008j.html#59 dollar coins
http://www.garlic.com/~lynn/2009c.html#7 Payments start-up Noca takes aim at interchange Achilles heel
http://www.garlic.com/~lynn/2009e.html#59 Tesco to open 30 "bank branches" this year
http://www.garlic.com/~lynn/2009f.html#60 Cobol hits 50 and keeps counting
http://www.garlic.com/~lynn/2009g.html#20 IBM forecasts 'new world order' for financial services
http://www.garlic.com/~lynn/2009g.html#57 LexisNexis says its data was used by fraudsters
http://www.garlic.com/~lynn/2009g.html#62 Solving password problems one at a time, Re: The password-reset paradox
http://www.garlic.com/~lynn/2009h.html#3 Consumer Credit Crunch and Banking Writeoffs
http://www.garlic.com/~lynn/2009h.html#19 Does anyone know of merchants who have successfully bypassed interchange costs
http://www.garlic.com/~lynn/2009h.html#78 Kansas City Fed Chief Espouses ACH for Debit Card Processing
http://www.garlic.com/~lynn/2009i.html#50 64 Cores -- IBM is showing a prototype already
http://www.garlic.com/~lynn/2009i.html#51 64 Cores -- IBM is showing a prototype already
http://www.garlic.com/~lynn/2009i.html#77 Financial Regulatory Reform - elimination of loophole allowing special purpose institutions outside Bank Holding Company (BHC) oversigh
http://www.garlic.com/~lynn/2009j.html#1 Is it possible to have an alternative payment system without riding on the Card Network platforms?
http://www.garlic.com/~lynn/2009j.html#50 How can we stop Credit card FRAUD?
http://www.garlic.com/~lynn/2009l.html#39 Network Rivalry Sparks 10-Year Quadrupling of PIN-Debit Pricing
http://www.garlic.com/~lynn/2009m.html#49 Hacker charges also an indictment on PCI, expert says
http://www.garlic.com/~lynn/2009m.html#62 August 7, 1944: today is the 65th Anniversary of the Birth of the Computer
http://www.garlic.com/~lynn/2009n.html#26 Signature specification without certificates
http://www.garlic.com/~lynn/2009p.html#68 US retailers face $100bn in ID fraud losses a year - study
http://www.garlic.com/~lynn/2009q.html#75 Now is time for banks to replace core system according to Accenture
http://www.garlic.com/~lynn/2010.html#70 Post Office bank account 'could help 1m poor'
http://www.garlic.com/~lynn/2010.html#98 Korean bank Moves back to Mainframes (...no, not back)

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

"The Naked Mainframe" (Forbes Security Article)

Refed: **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **
From: lynn@GARLIC.COM (Anne & Lynn Wheeler)
Subject: Re: "The Naked Mainframe" (Forbes Security Article)
Newsgroups: bit.listserv.ibm-main
Date: 22 Jan 2010 11:56:23 -0800
tony@HARMINC.NET (Tony Harminc) writes:
Right. And when VM had 20,000 licences, most of them were running PROFS on CMS. (Remember the Reagan White House, and Ollie North...)

recent post also referring to above:
http://www.garlic.com/~lynn/2010b.html#8 Happy DEC-10 Day
other recent PROFS reference:
http://www.garlic.com/~lynn/2010b.html#44 sysout using machine control instead of ANSI control

from the 60s, CMS was mainframe "personal computing" .... including some number of commercial online timesharing service bureaus dating back to 60s with cp67/cms (much more than email). tymshare had done their online computer conferencing on their vm-based commercial timesharing service ... and offerred it free to SHARE members (as VMSHARE) starting in aug76, archive:
http://vm.marist.edu/~vmshare/

one of the biggest such online operations was the world-wide internal (vm-based) HONE system ... eventually all branch office people in the world; not long after introduction of HONE ... it became requirement that ALL mainframe orders be processed via HONE applications
http://www.garlic.com/~lynn/subtopic.html#hone

then mid-range price/performance dropped below some threshold and 43xx saw gigantic explosion starting in the late 70s ... similar to what DEC saw with vax/vms.

old post with decade of vax/vms numbers sliced and diced by year, model, US & non-us:
http://www.garlic.com/~lynn/2002f.html#0 Computers in Science Fiction

a big differentiator between 43xx and vax/vms ... were some large commercial customers with orders of multiple hundreds at a time (the smaller order sizes were otherwise similar)

the change in the mid-80s was workstations and large PCs were starting to take over that mid-range computer market (and PCs starting to subsume CMS personal computing). the continued large volumes that endicott expected to see for the 43xx follow-ons never materialized.

misc. old 43xx-related email from the period
http://www.garlic.com/~lynn/lhwemail.html#43xx

at one point, somebody from pok gave a talk in san fran ... and made some statement about 11,000 of the vax sales should have been 43xx (would have been good size shift ... see numbers in above post) ... because 43xx provided better price/performance.

however, it wasn't just dec/vax that 43xx was affecting. old email with references to 4341, 158, & 3031 benchmarks:
http://www.garlic.com/~lynn/2006y.html#email790212b
in this post
http://www.garlic.com/~lynn/2006y.html#21 moving on

also, customers were finding that a vm/4341 cluster was cheaper than 3033, higher aggregate mip rate, much larger aggregate storage, and higher aggregate i/o capacity. There is folklore, that because of the above ... at one point, POK directed Fishkill to cut the Endicott allocation in half for a critical component needed for 4341 manufacturing.

One of the things that was happening by the mid-70s ... as processing power was increasing ... disk thruput improvements weren't keeping pace with processor speed improvements. as a result, systems were having to rely more & more on larger & larger electronic storage ... to compensate for the growing disk i/o bottleneck. 370s were stuck with 24bit addressing and 16mbyte virtual and real storage ... which resulted in significant constrained operation for many 3033s.

3033 eventually came up with a hack for >16mbyte real storage, using IDALs and slight-of-hand with two unused bits in the PTE ... although there was still an issue with some things having to be "below the line". One of the issues was that part of the solution involved virtual pages that were above the line having to be moved below the line ... and they were going to rely on IDALs to write it out to disk and then read it back in (below the line). Old email referring to hack I gave them to do the move w/o having to do I/O:
http://www.garlic.com/~lynn/2006t.html#email800121
in this post
http://www.garlic.com/~lynn/2006t.html#15 more than 16mbyte support for 370

misc other "below the line" posts
http://www.garlic.com/~lynn/95.html#2 Why is there only VM/370?
http://www.garlic.com/~lynn/2000d.html#82 "all-out" vs less aggressive designs (was: Re: 36 to 32 bit transition)
http://www.garlic.com/~lynn/2003f.html#4 Alpha performance, why?
http://www.garlic.com/~lynn/2004f.html#38 Infiniband - practicalities for small clusters
http://www.garlic.com/~lynn/2005.html#34 increasing addressable memory via paged memory?
http://www.garlic.com/~lynn/2005p.html#19 address space
http://www.garlic.com/~lynn/2005q.html#30 HASP/ASP JES/JES2/JES3
http://www.garlic.com/~lynn/2005u.html#44 POWER6 on zSeries?
http://www.garlic.com/~lynn/2006l.html#2 virtual memory
http://www.garlic.com/~lynn/2006w.html#23 Multiple mappings
http://www.garlic.com/~lynn/2007b.html#34 Just another example of mainframe costs
http://www.garlic.com/~lynn/2008f.html#12 Fantasy-Land_Hierarchal_NUMA_Memory-Model_on_Vertical
http://www.garlic.com/~lynn/2009g.html#71 308x Processors - was "Mainframe articles"
http://www.garlic.com/~lynn/2009g.html#74 308x Processors - was "Mainframe articles"
http://www.garlic.com/~lynn/2010.html#84 locate mode, was Happy DEC-10 Day

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

"The Naked Mainframe" (Forbes Security Article)

Refed: **, - **, - **, - **, - **
From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: "The Naked Mainframe" (Forbes Security Article)
Newsgroups: bit.listserv.ibm-main
Date: Fri, 22 Jan 2010 15:13:12 -0500
R.Skorupka@BREMULTIBANK.COM.PL (R.S.) writes:
I would like to see the numbers above with avg MIPS for every kind of OS. I guess that MIPS number is much higher for TPF and z/OS and quite low for VSE. Last but not least: I do not expect to much (any?) customer with z/VM and no other OS. Including zLinux.

recent thread discussing rewrite of major TPF application that was projected could grow to several hundred ES9000 ... but with the rewrite being able to handle the growth on ten RS/6000 580s.
http://www.garlic.com/~lynn/2010b.html#80 Happy DEC-10 Day

earlier pieces of the thread:
http://www.garlic.com/~lynn/2010b.html#74 Happy DEC-10 Day
http://www.garlic.com/~lynn/2010b.html#78 Happy DEC-10 Day
http://www.garlic.com/~lynn/2010b.html#79 Happy DEC-10 Day

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Remember Ed Curry!

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Remember Ed Curry!
Newsgroups: alt.folklore.computers
Date: Fri, 22 Jan 2010 15:24:28 -0500
Chris Barts <chbarts+usenet@gmail.com> writes:
Who Hides the Truth About NT, and Why?:
http://cryptome.org/ed-curry.htm
Date: Sun, 12 Sep 1999 18:16:01 +0400
...
Ed Curry is formerly a military man, a NSA-certified technical security analyst, and a former independent contractor for the Microsoft Corporation, who for several years has been saying that the Pentagon and other US government agencies have violated their own security rules by purchasing mass quantities of a non-secure computer operating system, Windows NT.
...
It's getting hard to find information about Curry online and I'm pretty sure it's impossible to find it anywhere else by now. His investigation and complaints didn't go anywhere, after all. This just demonstrates how right he was, even if he didn't see this specific threat at the time.


re:
http://www.garlic.com/~lynn/2010b.html#75 Windows plagued by 17-year-old privilege escalation bug

and what has been the justification for the move to common criteria?

I was at a presentation a few years ago that of something like 60+ EAL evaluations ... all but 2 or 3 had unpublished (secret?) deviations from standard protection profile.

in the past, I've been hassled by some from the common criteria camp about still carrying orange book stuff in my merged security taxonomy and glossary
http://www.garlic.com/~lynn/index.html#glosnote

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Happy DEC-10 Day

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Happy DEC-10 Day
Newsgroups: alt.sys.pdp10, alt.folklore.computers
Date: Fri, 22 Jan 2010 15:52:08 -0500
Morten Reistad <first@last.name> writes:
This is the point about "developing the Internet". We need an interactive, public scripting language that can exist in a validated sandbox, and that can interace, full-duplex, with the user _and_ with the service provider. The clues here are public script language, interactive, full-duplex and a sandbox that also includes a server. These four attributes make this very distinct from an http/https based service.

some drift to my wife's arguments with SNA group when she was con'ed into going to POK to be in charge of mainfame loosely-coupled (cluster) architecture ... and SNA group constantly insisting that she had to use (half-duplex) SNA for loosely-coupled coordination. There was (temporary) truce that she could use whatever she wanted within the boundaries of of the datacenter walls ... but SNA still had to be used for anything that crossed the datacenter boundary.

a couple years later, research was doing vm/4341 cluster operation and was using a full-duplex broadcast protocol for cluster coordination (for various cluster-wide operations, it was small subsecond elapsed time). However, to ship vm/4341 cluster support to customers ... they were forced to move to (half-duplex) SNA ... and things that had been taking small subsecond elapsed time were all of sudden taking greater than 30 seconds (even tho the hardware was still identical).

misc. past posts mentioning her peer-coupled shared-data architecture
http://www.garlic.com/~lynn/submain.html#shareddata

the stand requiring SNA to be used for everything that crossed the boundaries of the datacenter walls ... played significant role in the whole terminal emulation period
http://www.garlic.com/~lynn/subnetwork.html#emulation

and fighting off things like client/server and 3-tier architecture
http://www.garlic.com/~lynn/subnetwork.html#3tier

--
40+yrs <b>virtualization</b> experience (since Jan68), online at home since Mar1970

Happy DEC-10 Day

Refed: **, - **, - **
From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Happy DEC-10 Day
Newsgroups: alt.sys.pdp10, alt.folklore.computers
Date: Fri, 22 Jan 2010 16:32:18 -0500
Eric Chomko <pne.chomko@comcast.net> writes:
Clearly having microcomputers in the business world was significant, but it would NEVER become what it is today unless people bought them for their homes as well. The reason is that the advent of PCs in the home drove down the price due to an increase in demand. Sure IBM would have loved charging more for their PCs and kept them out of homes and at the office, but people would have a choice to have a home computer with business apps and tell IBM what they could do with their high- priced computers. The clone market sort of headed that whole thing off, and traditional home computers started to add business software to their computers as well.

i would claim that there was chicken&egg ... the hobbiest stuff had difficulty in the home market. the volumes that PC got from business market ... especially since it was about the same price as 3270 and businesses buying tens of thousands of 3270s could get PC for about the same price and do both 3270 and some local computing (negligible incremental business justification for PC if 3270 terminal was already justified). that resulted a big jump in volume ... and attracted a lot of developers (because of the volume/install base).

then things snowballed ... volumes got more developers, more development got more volumes ... volumes got attention of clones, clones&volumes further reduced price, reduction in price increased volumes, reduced priced allowed moving into more markets, further increasing volumes & reducing price.

in the early 80s, my brother was regional sales for apple (claimed largest physical region in conus) and would periodically come to town ... and I would be invited to after-work business dinners. I got to argue with some mac developers (before it was announced) whether or not mac needed terminal/3270-emulation feature in order to get the volumes going.

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Who's to Blame for the Meltdown?

Refed: **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **
From: lynn@garlic.com (Lynn Wheeler)
Date: 22 Jan, 2010
Subject: Who's to Blame for the Meltdown?
Blog: Greater IBM
a few posts in various meltdown threads over the past couple years:

http://www.garlic.com/~lynn/2008c.html#76 Neglected IT Tasks May Have Led to Bank Meltdown
http://www.garlic.com/~lynn/2008o.html#80 Can we blame one person for the financial meltdown?
http://www.garlic.com/~lynn/2008q.html#49 Have not the following principles been practically disproven, once and for all, by the current global financial meltdown?
http://www.garlic.com/~lynn/2008s.html#5 Greed - If greed was the cause of the global meltdown then why does the biz community appoint those who so easily succumb to its temptations?
http://www.garlic.com/~lynn/2009b.html#73 What can we learn from the meltdown?
http://www.garlic.com/~lynn/2009c.html#39 'WHO IS RESPONSIBLE FOR THE GLOBAL MELTDOWN'

time had series on it:

25 People to Blame for the Financial Crisis; Phil Gramm
http://content.time.com/time/specials/packages/article/0,28804,1877351_1877350_1877330,00.html

most recent thread that wandered off into the subject:
http://www.garlic.com/~lynn/2010b.html#82
http://www.garlic.com/~lynn/2010b.html#86

search engine history, was Happy DEC-10 Day

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: search engine history, was Happy DEC-10 Day
Newsgroups: alt.sys.pdp10, alt.folklore.computers
Date: Fri, 22 Jan 2010 18:29:56 -0500
Jim Stewart <jstewart@jkmicro.com> writes:
MIPS processors were used in Tivo series 1 and 2, running Linux. One of the first mainstreet embedded Linux products and one that exposed a bunch of casual hackers to the architecture.

there used to be jokes about their actually only were 200 people in the industry
http://www.garlic.com/~lynn/2010b.html#67 How long for IBM System/360 architecture and its descendants?

mentions that the executive we reporting to when we were doing our HA/CMP product
http://www.garlic.com/~lynn/subtopic.html#hacmp

then moved over to head-up somerset (motorola, ibm, apple, et al to do power/pc); after a stint there ... then about the time we went back to san jose ... he took the job as president of mips. all the executives got a personal indy ... so I offerred to order it for him and take it home and configure it ... it stayed home until he left that job (and I had to turn it back in)

misc. past posts mentioning 801, risc, iliad, romp, rios, power, power/pc, somerset, etc
http://www.garlic.com/~lynn/subtopic.html#801

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Oldest Instruction Set still in daily use?

Refed: **, - **, - **, - **
From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Oldest Instruction Set still in daily use?
Newsgroups: alt.folklore.computers
Date: Sat, 23 Jan 2010 08:34:34 -0500
jmfbahciv <jmfbahciv@aol> writes:
yes, Hiding. There was a big stink when Pelosi insisted that the meetings to iron out the differences between the two bills be done in closed sessions. Nobody knows what that new bill looks like. It's also not in the news because nobody has anything to report.

re:
http://www.garlic.com/~lynn/2010b.html#82 Oldest Instruction Set still in daily use?

in the case of Gramm-Leach-Bliley, it included adding provisions to go from 54-44 vote to 90-8 ... aka making it "veto proof"
http://en.wikipedia.org/wiki/Gramm-Leach-Bliley_Act

the above mentions adding provisions for financial privacy ... but the "opt-out" flavor was viewed as "federal pre-emption" of pending "opt-in" in progress in cal.

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

Oldest Instruction Set still in daily use?

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Oldest Instruction Set still in daily use?
Newsgroups: alt.folklore.computers
Date: Sat, 23 Jan 2010 08:49:20 -0500
Anne & Lynn Wheeler <lynn@garlic.com> writes:
the above mentions adding provisions for financial privacy ... but the "opt-out" flavor was viewed as "federal pre-emption" of pending "opt-in" in progress in cal.

re:
http://www.garlic.com/~lynn/2010b.html#82 Oldest Instruction Set still in daily use?
http://www.garlic.com/~lynn/2010b.html#94 Oldest Instruction Set still in daily use?

recent article with different kind of "opt-out" vis-a-vis "opt-in"

Spammer trick: exploiting CAN-SPAM loopholes
http://blogs.computerworld.com/15418/spammer_trick_exploiting_can_spam_loopholes

from above:

Ultimately, this is the downside of spam laws that codify an opt-out regime. As I noted in November, most of the rest of the world requires that marketers first get a user's permission. The gold standard laws are the ones that also specify the permission be 'informed' -- i.e., the user's not being tricked into giving permission and has sufficient information to make a choice.

... snip ...

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

"The Naked Mainframe" (Forbes Security Article)

Refed: **, - **, - **, - **, - **, - **, - **, - **, - **, - **, - **
From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: "The Naked Mainframe" (Forbes Security Article)
Newsgroups: bit.listserv.ibm-main
Date: Sat, 23 Jan 2010 09:33:36 -0500
ps2os2@YAHOO.COM (Ed Gould) writes:
Now this goes back aways and my memory is not 100 percent but...

There are probably some of you here that remember when the White House (in the 70's-80's) lost a lot of email from around the time of Watergate.

I had a friend who was an IBMer working at the White House in that time frame. He was involved in trying to get back all the lost emails. My memory is iffy here but when I was talking with my friend he was telling me how exhaustive IBM worked at getting back the emails. They had quite a few factory types working on getting them back. They were not particularly successful because of data getting written over and recovery was at best spotty.

I am pretty sure they ran some type of PROFs (it was VM based that is all I remember) and he got somewhat familiar with reading track dumps and also whiz bang (he never gave me specifics) way of reading what was underneath what was current on the track.

He is long time retired and is enjoying a well deserved retirement so I hope he doesn't get in any trouble for anything I am writing here.


re:
http://www.garlic.com/~lynn/2010b.html#8 Happy DEC-10 Day
http://www.garlic.com/~lynn/2010b.html#44 sysout using machine control instead of ANSI control
http://www.garlic.com/~lynn/2010b.html#87 "The Naked Mainframe" (Forbes Security Article)

folklore is that one would need one large pile of clearances to touch such stuff (sufficient to cover anything that might be on the media).

tape management (including backup tapes) doesn't always get the consideration it deserves ... recent posts about almaden datacenter going thru a period where it appeared scratch tape requests involved selecting tapes at random to mount:
http://www.garlic.com/~lynn/2010b.html#51 Source code for s/360

there have been past stories about some companies nearly being taken down ... when it was discovered that backup process wasn't actually writing anything on the tapes.

there are some federal standards for overwriting as countermeasure to such recovery

some discussion
http://en.wikipedia.org/wiki/Data_remanence
http://en.wikipedia.org/wiki/Data_recovery
http://en.wikipedia.org/wiki/Data_erasure
http://en.wikipedia.org/wiki/Data_Shredder

one of the Rainbow books:

A Guide to Understanding Data Remanence in Automated Information Systems
http://www.fas.org/irp/nsa/rainbow/tg025-2.htm

from above:
5.2.1 MAGNETIC TAPES

Although overwriting can be used for clearing this media, the method is time consuming and generally never used. Also, inter-record gaps may preclude proper clearing. A better method for clearing Type I, II, and III tapes is degaussing with a Type I or Type II degausser. This procedure is considered acceptable for clearing, but not purging, all types of tapes.

Degaussing with an appropriate degausser is the only method the DoD accepts for purging this media. Specifically, a Type I degausser can purge only Type I tapes, and Type II degaussers can purge Types I and II tapes. No degausser presently exists that is capable of purging Type III tapes in accordance with NSA/CSS Specification L14-4-A.


... snip ...

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

"The Naked Mainframe" (Forbes Security Article)

Refed: **, - **, - **, - **, - **, - **, - **, - **, - **, - **
From: lynn@GARLIC.COM (Anne & Lynn Wheeler)
Subject: Re: "The Naked Mainframe" (Forbes Security Article)
Newsgroups: bit.listserv.ibm-main
Date: 23 Jan 2010 11:16:16 -0800
R.Skorupka@BREMULTIBANK.COM.PL (R.S.) writes:
Well... I remember Ronald Reagan, I know what White House is, but neither I know Ollie North, nor I know what is PROFS and what is relationship of VM to Ronald Reagan.

re:
http://www.garlic.com/~lynn/2010b.html#87 "The Naked Mainframe" (Forbes Security Article)
http://www.garlic.com/~lynn/2010b.html#88 "The Naked Mainframe" (Forbes Security Article)
http://www.garlic.com/~lynn/2010b.html#96 "The Naked Mainframe" (Forbes Security Article)

vm systems all over the federal gov. ... and (VM) PROFS email in several gov. agencies ... including the executive branch. other recent posts mentioning PROFS
http://www.garlic.com/~lynn/2010b.html#8 Happy DEC-10 Day
http://www.garlic.com/~lynn/2010b.html#44 sysout using machine control instead of ANSI control

Ollie North reference
http://en.wikipedia.org/wiki/Oliver_North

there was congressional investigation and supposedly some amount of evidence came from email archive/backup.
http://en.wikipedia.org/wiki/Iran-Contra_affair

slightly related
http://www.garlic.com/~lynn/2010b.html#63 Source code for s/360 [PUBLIC]

reference to virtual machine systems back to 60s & cp67:
http://web.archive.org/web/20090117083033/http://www.nsa.gov/research/selinux/list-archive/0409/8362.shtml

later example
http://www.garlic.com/~lynn/2001m.html#email790404b
in this post
http://www.garlic.com/~lynn/2001m.html#15 departmental servers

from various 43xx email
http://www.garlic.com/~lynn/lhwemail.html#43xx

where possible AFDS twenty vm/4341 order grew to 210.

Originally mentioned old email
http://www.garlic.com/~lynn/2001m.html#email790404
in
http://www.garlic.com/~lynn/2001m.html#12 Multics Nostalgia

which was small needle regarding AFDS & having been an early Multics strong-hold. Some of the CTSS people had gone to the 5th flr and Multics; others had gone to the science center on the 4th flr and did virtual machine systems (initially cp40, which morphed into cp67 and later vm370). Multics sites
http://www.multicians.org/sites.html
AFDS
http://www.multicians.org/site-afdsc.html

for other drift ... reference to an IBM paper:
http://www.garlic.com/~lynn/2002l.html#42 Thirty Years Later: Lessons from the Multics Security Evaluation

originally here ... but since gone 404
domino.watson.ibm.com/library/cyberdig.nsf/papers/FDEFBEBC9DD3E35485256C2C004B0F0D/$File/RC22534.pdf

Multics reference to above as well as the original study:
http://www.multicians.org/security.html

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

"The Naked Mainframe" (Forbes Security Article)

Refed: **, - **, - **, - **
From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: "The Naked Mainframe" (Forbes Security Article)
Newsgroups: bit.listserv.ibm-main
Date: Sat, 23 Jan 2010 16:53:59 -0500
lynn@GARLIC.COM (Anne & Lynn Wheeler) writes:
Originally mentioned old email
http://www.garlic.com/~lynn/2001m.html#email790404
in
http://www.garlic.com/~lynn/2001m.html#12 Multics Nostalgia


re:
http://www.garlic.com/~lynn/2010b.html#87 "The Naked Mainframe" (Forbes Security Article)
http://www.garlic.com/~lynn/2010b.html#88 "The Naked Mainframe" (Forbes Security Article)
http://www.garlic.com/~lynn/2010b.html#96 "The Naked Mainframe" (Forbes Security Article)
http://www.garlic.com/~lynn/2010b.html#97 "The Naked Mainframe" (Forbes Security Article)

above example is branch office person using the (vm-based, world-wide sales&marketing) HONE system to send me email ... misc. past HONE references:
http://www.garlic.com/~lynn/subtopic.html#hone

providing highly enhanced systems to internal locations was one of my hobbies ... and HONE was regular customer.

I previously mentioned VMSHARE ... online computer conferencing system offered to SHARE (starting aug76)
http://vm.marist.edu/~vmshare/

not a lot of employees got direct access to VMSHARE ... so I set up processes that TYMSHARE would regularly send me complete dump of all VMSHARE (and later, PCSHARE) files ... and I would put them up on several internal systems, including HONE. Another example, somebody from branch office in Kuwait sending me email (regarding VMSHARE info)
http://www.garlic.com/~lynn/2007b.html#email830227

other internal email (from Paris) ... this time about PCSHARE
http://www.garlic.com/~lynn/2007b.html#email821214
http://www.garlic.com/~lynn/2007b.html#email821217

Paris-La Defense was one of the first overseas system installations I was involved in, when EMEA hdqtrs moved from the states in the early 70s (back then, overseas links weren't so pervasive, so it was lot harder to figure out how to logon back home to read email).

The above was small part of what got me blamed for computer conferencing on the internal network in the late 70s and early 80s ... misc. past posts mentioning internal network
http://www.garlic.com/~lynn/subnetwork.html#internalnet

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

"The Naked Mainframe" (Forbes Security Article)

Refed: **, - **, - **, - **, - **, - **, - **
From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: "The Naked Mainframe" (Forbes Security Article)
To: <ibm-main@bama.ua.edu>
Date: Sat, 23 Jan 2010 17:37:19 -0500
bbreynolds <bbreynolds@aol.com> writes:
Was that a component which was shared by the 3033? Something unique to the the 4341? I know that IBM's internal politics were sometimes off the wall, but that folklore seems extreme.

re:
http://www.garlic.com/~lynn/2010b.html#87 "The Naked Mainframe" (Forbes Security Article)
http://www.garlic.com/~lynn/2010b.html#88 "The Naked Mainframe" (Forbes Security Article)
http://www.garlic.com/~lynn/2010b.html#96 "The Naked Mainframe" (Forbes Security Article)
http://www.garlic.com/~lynn/2010b.html#97 "The Naked Mainframe" (Forbes Security Article)
http://www.garlic.com/~lynn/2010b.html#98 "The Naked Mainframe" (Forbes Security Article)

Really long email ... heavily edited. Mostly trip report about extended east coast trip hitting several places.

Date: 04/29/79 16:39:03
From: wheeler

<< lots & lots of stuff snipped >>

While I was in Endicott, I think I talked them into putting me on the distribution list for VM functional spec. documents. After the visit to Endicott, we went by Cornell Univ. for an afternoon/evening. They had a number of interesting things to say. We have talked before about doing a joint study with them on their mini-disk manager. They finally asked xxxxx about it at the last Share meeting. He hemmed and hawed around for a long time not sounding very hopeful and finally said any such undertaking has to be approved by YYYYY. MIT Prof was also there giving a seminar for a week or 2. They had a funny story to tell. On the 1st day MIT Prof had some not very complimentary things to say about Cornell's comp. science department. They took him aside at lunch and told him that wasn't exactly the correct thing to do. He apparently held his tongue for a whole week. Finally he had the opportunity to state that if all computers at Cornell were destroyed the computer science department would never know about it.

After Cornell we went by Kingston and then POK. In both Endicott and POK had some very interesting discussions about confidential stuff that is going on. In Endicott especially, there was even a hardware modification design session which I think we work some stuff out. Finally found out what head-of-POK was going to do about the 4341. I all along thot he would force Endicott into slowing the machine down. I guess he couldn't come with a way. He did come up with something that is probably even more effective tho. He somehow arraigned for the East Fishkill plant to cut their hardware output allocation to Endicott in half. There were comments that head-of-POK was called several choice names. Endicott still may win tho.

<< lots & lots of stuff snipped >>


... snip ... top of post, old email index

Had pretty close working relationship with Cornell over the years, for other drift when we were ramping up to do the NSFNET backbone (before internal politics shut us down) ... Cornell was one of the players; old email reference
http://www.garlic.com/~lynn/2006u.html#email860505
in this post
http://www.garlic.com/~lynn/2006u.html#56 Ranking of non-IBM mainframe builders?

misc. old NSFNET-related email
http://www.garlic.com/~lynn/lhwemail.html#nsfnet

tcp/ip is the technology basis for the modern internet but NSFNET backbone was the operational basis for the modern internet (and CIX was the business basis for the modern internet).

The director of NSF attempted to help out writing a letter to couple people in the corporation (copying the CEO), but that just aggravated the internal politics ... recent reference:
http://www.garlic.com/~lynn/2009q.html#42 The 50th Anniversary of the Legendary IBM 1401

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970

"The Naked Mainframe" (Forbes Security Article)

Refed: **, - **, - **, - **, - **, - **, - **, - **, - **
From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: "The Naked Mainframe" (Forbes Security Article)
Newsgroups: bit.listserv.ibm-main, alt.folklore.computers
Date: Sun, 24 Jan 2010 10:20:29 -0500
there seems to be some hiccup with this recent post (I did twice) between the mailing list and usenet (missing on usenet, but I finally checked mailing list archive; I normally read on usenet, but post to the mailing list).
http://www.garlic.com/~lynn/2010b.html#99 "The Naked Mainframe" (Forbes Security Article)

also small ending piece snipped from previous email
http://www.garlic.com/~lynn/2010b.html#email790429

Date: 04/29/79 16:39:03
From: wheeler

<< lots & lots of stuff snipped >>

One more thing about Endicott, their datacenter production VM system is so backlevel, they asked about SJRL's VM system. There were tentative plans made for some Endicott people to come out to SJRL and pick up our floor system for installation in Endicott. Some of this in light of the hardware error recovery that we have been adding, especially in response to the problems in the DASD engineering labs but also to normal problems we have here.


... snip ... top of post, old email index

old email (year later) about sjr/vm distribution
http://www.garlic.com/~lynn/2006u.html#email800429
in this post (also contains several other old email pieces)
http://www.garlic.com/~lynn/2006u.html#26 Assembler question

other past references to SJR/VM system
http://www.garlic.com/~lynn/2007b.html#51 Special characters in passwords was Re: RACF - Password rules
http://www.garlic.com/~lynn/2007c.html#12 Special characters in passwords was Re: RACF - Password rules
http://www.garlic.com/~lynn/2008h.html#46 Whitehouse Emails Were Lost Due to "Upgrade"
http://www.garlic.com/~lynn/2008s.html#39 The Internet's 100 Oldest Dot-Com Domains
http://www.garlic.com/~lynn/2009i.html#35 SEs & History Lessons

they let me play engineer over in bldgs. 14&15 ... some past posts
http://www.garlic.com/~lynn/subtopic.html#disk

they had been running (pre-scheduled) stand-alone testing of engineering/development hardware (several dedicated, stand-alone 370s). At one time they had tried to use MVS in the environment, but had experienced 15min MTBF. I undertook to completely rewrite i/o supervisor so that it would never fail and they could concurrently test several devices (on demand, instead of the around-the-clock pre-scheduled, stand-alone testing that they had been doing).

Mentioning the 15min MTBF in a purely internal-only report, brought the wrath of the MVS group down on my head ... but seems small in comparison to the effort to cut allocation to endicott for building 4341s.

in any case, endicott datacenter people never came out, i think somebody got around to how would it look if endicott datacenter was running one of my vm systems.

I've mentioned before that after demise to future system,
http://www.garlic.com/~lynn/submain.html#futuresys

there was mad rush to get stuff (hardware & software) back into the 370 product pipeline ... as well as getting around to kicking off XA effort (eventually "811" from the date on the hardware architecture documents). POK also made the case to corporate that in order to make the mvs/xa ship schedule, had to kill vm370, shutdown the vm370 development group and move all the people to POK ... a couple recent posts:
http://www.garlic.com/~lynn/2010b.html#37 Happy DEC-10 Day
http://www.garlic.com/~lynn/2010b.html#44 sysout using machine control instead of ANSI control

Endicott eventually managed to save the vm370 product mission (seeing the leading edge of what was to become the vm midrange explosion), but had to reconstitute a group from scratch. By the time of the above email, they were still ramping up.

misc. other posts in this thread:
http://www.garlic.com/~lynn/2010b.html#87 "The Naked Mainframe" (Forbes Security Article)
http://www.garlic.com/~lynn/2010b.html#88 "The Naked Mainframe" (Forbes Security Article)
http://www.garlic.com/~lynn/2010b.html#96 "The Naked Mainframe" (Forbes Security Article)
http://www.garlic.com/~lynn/2010b.html#97 "The Naked Mainframe" (Forbes Security Article)
http://www.garlic.com/~lynn/2010b.html#98 "The Naked Mainframe" (Forbes Security Article)

--
40+yrs virtualization experience (since Jan68), online at home since Mar1970


previous, next, index - home