List of Archived Posts

2002 Newsgroup Postings (04/27 - 05/20)

Blade architectures

WATFOR's Silver Anniversary

Computers in Science Fiction

Why are Mainframe Computers really still in use at all?

markup vs wysiwyg (was: Re: learning how to use a computer)

Black magic in POWER5

The HTML one-pixel dilation trick

income tax [was: Computers in Science Fiction]

The HTML one-pixel dilation trick

IBM MIcrochannel??

"Soul of a New Machine" Computer?

"Soul of a New Machine" Computer?

"Soul of a New Machine" Computer?

Secure Device Drivers

"Soul of a New Machine" Computer?

Security Issues of using Internet Banking

Why are Mainframe Computers really still in use at all?

Black magic in POWER5

Black magic in POWER5

Why did OSI fail compared with TCP-IP?

Security Issues of using Internet Banking

Why did OSI fail compared with TCP-IP?

Why did OSI fail compared with TCP-IP?

Why are Mainframe Computers really still in use at all?

Why did OSI fail compared with TCP-IP?

Why are Mainframe Computers really still in use at all?

Why did OSI fail compared with TCP-IP?

Security Issues of using Internet Banking

Why did OSI fail compared with TCP-IP?

Why did OSI fail compared with TCP-IP?

Why did OSI fail compared with TCP-IP?

Why did OSI fail compared with TCP-IP?

Secure Device Drivers

ESCON Distance Limitations - Why ?

Why did OSI fail compared with TCP-IP?

Why did OSI fail compared with TCP-IP?

The need for Meaningful Error Messages :)

Security Issues of using Internet Banking

Why is DSA so complicated?

"Soul of a New Machine" Computer?

Why did OSI fail compared with TCP-IP?

Why is DSA so complicated?

Why is DSA so complicated?

Why is DSA so complicated?

ibm icecube -- return of watercooling?

Why did OSI fail compared with TCP-IP?

Why did OSI fail compared with TCP-IP?

Why are Mainframe Computers really still in use at all?

Why did OSI fail compared with TCP-IP?

Why did OSI fail compared with TCP-IP?

Why did OSI fail compared with TCP-IP?

Timecard stupidity

Spotting BAH Claims to Fame

Are you sure about MONDEX?

Are you sure about MONDEX?

Multics hardware (was Re: "Soul of a New Machine" Computer?)

Siemens ID Device SDK (fingerprint biometrics) ???

Amiga Rexx

Amiga Rexx

Amiga Rexx

Amiga Rexx

GE 625/635 Reference + Smart Hardware

ibm icecube -- return of watercooling?

Pardon my ignorance,

Pardon my ignorance,

Real man-in-the-middle attacks?

Formal Classification for Security Topics

Coulda, Woulda, Shoudda moments?

Why are Mainframe Computers really still in use at all?

Digital signature

Pipelining in the past

Coulda, Woulda, Shoudda moments?

Biometrics not yet good enough?

Coulda, Woulda, Shoudda moments?

Coulda, Woulda, Shoudda moments?

Coulda, Woulda, Shoudda moments?

Pipelining in the past

Pipelining in the past

Is it safe to use social securty number as intranet username?

Coulda, Woulda, Shoudda moments?

Pipelining in the past

Multics reference in Letter to Editor

Future architecture

Questions about computer security

Questions on IBM Model 1630

Blade architectures

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Blade architectures
Newsgroups: alt.folklore.computers
Date: Sat, 27 Apr 2002 18:14:24 GMT

Kragen Sitaker writes:

And, of course, KeyKOS was a single-level-store OS for the 370 and
SPARC, which I seem to recall was somewhat of a commercial success in
the 1980s.  It wasn't the case that everything had an address in
KeyKOS, though; although you could map any data into your address
space if you had the proper access to it (just as you can in Linux
with mmap()), that wasn't the normal way of operating.

others for the 370 was RASP (simpson worked on before joining
Amdahl) and Aspen (after he joined amdahl). misc. refs
http://www.garlic.com/~lynn/2000f.html#68 TSS ancient history, was X86 ultimate CISC? designs)
http://www.garlic.com/~lynn/2000f.html#69 TSS ancient history, was X86 ultimate CISC? designs)
http://www.garlic.com/~lynn/2000f.html#70 TSS ancient history, was X86 ultimate CISC? designs)
http://www.garlic.com/~lynn/2001b.html#73 7090 vs. 7094 etc.

--
Anne & Lynn Wheeler   | lynn@garlic.com -  http://www.garlic.com/~lynn/

WATFOR's Silver Anniversary

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: WATFOR's Silver Anniversary
Newsgroups: bit.listserv.ibm-main,alt.folklore.computers
Date: Sun, 28 Apr 2002 15:09:02 GMT

nospam@nowhere.com   (Steve Myers) writes:

The big advantage of both WATFOR systems was their run time
diagnostics.  References to undefined variables, and array bounds
checking, were godsends to both students and professionals.

Many people that ran WATFOR/360 remember those crazy error codes,
though, and it was just as bad with WATFOR/7040.  That was the
biggest single problem with both systems.

typical student job was under second compile & run under watfor (on
360/65) ... one of the big wins for students with watfor runtime was
things like uninitialized variables & bounds checking.

benchmark i presented at fall '68 SHARE in boston:
http://www.garlic.com/~lynn/94.html#17

standard gen'ed system in above took approx. 30seconds elapsed time to
do 3step job ... essentially all job scheduler ... or about 10 seconds
per job step. even with the r17 loader as standard ... with a single
compile,linkedit, & go per student job that would still be around
11-12 seconds plus per student job.

fortran g compile was something like 10 times slower than than watfor
(but execution was faster). majoritqy of student jobs were compile ...
first a very large percentage didn't even make it to run-time because
of compile-time errors, second ... the watfor bounds check &
uninitialized variables terminated the program very early. even with a
smart loader handling multiple student jobs per job scheduler pass ...
there was still extensive disk i/o reloading fortran g compile (at
least half dozen different pds members) and link-edit accessing the
fortran run-time library (which would have to be done per every
student job).

watfor job step would load the compiler/runtime into memory in the
initial 10 second job step ... and then blow threw 100 student jobs in
a couple seconds (15 seconds or less total elapsed time, including job
scheduler & watfor compiler/runtime).

with the optimized sysgen that I showed in the above ... job scheduler
and other elapsed time processing was speeded up by over a factor of
over two times ... 12 seconds for three step job (instead of
30seconds) ... which resulted in around 4 seconds for watfor compiler
& runtime load and then blow thru 100 student jobs in less than
additional five seconds ... say 8 to 9 seconds elapsed time for 100
typical student jobs. It wasn't until you got to some of the higher
level courses that runtime elapsed time came close to exceeding
compile elapsed time ... at that point you saw some cross-over to
fortran g.

the problem with optimized sysgen was that normal PTF activity would
update/replace PDS member ... aka at the end of allocated PDS space
allocation ... rather at carefully positioned. 4-5 months of normal
PTF activity would increase the carefully optimized 12 seconds to 20
seconds elapsed time or more for 3step job scheduler processing.

later fortran H came along that was a really heavy weight compile time
...  but saw some significant runtime reductions for big
programs. then there was fortran Q ... internally, which was an
enhanced fortran H ...  done by the guy at PASC also responsible for
the APL\CMS microcode on the 370/145 (and some number of other
projects). fortran Q enhancements were eventually released as fortran
HX product.

--
Anne & Lynn Wheeler   | lynn@garlic.com -  http://www.garlic.com/~lynn/

Computers in Science Fiction

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Computers in Science Fiction
Newsgroups: alt.folklore.computers
Date: Sun, 28 Apr 2002 17:27:51 GMT

jmfbahciv writes:

Oh, no.  That's not how I see it.  I gave a talk about my USAGE
project to a customer one time.  He wanted to know if the data
capture was real time.  My answer, of course, was no.  He got
real upset (much to the consternation of the salesman sitting
in the meeting).  However, the saleman finally figured out that
the customer and I had completely different definitions of
realtime.  My definition had to do with the realtime clock
that could be an option on the CPU.  The customer's definition
was based in IBMese.  I never did find out what that was because
the salesman said that he'ld explain once I was out of the room...

Hint...Lynn? :-)

note that starting with 370 in the early 70s there was real-time TOD
clock that was 64 bits, the 33nd bit was just slightly over a second
(epoch slightly over 130 years ... originally epoch was suppose to
start first day of the last century ... which carried well into this
century, but the epoch is 1/1/70 ... so it carries it to the next
century). Most machines tic'ed the clock in real time at about a
microsecond (although the resolution provided for nanosecond) and the
store clock instruction was non-supervisor/non-priviledge instruction.
There was a supervisor instruction that could set a (clock comparator)
value that would interrupt when the clock reached that value. There
was also a separate 64bit interval timer that decrement at the same
rate as the TOD clock which would also interrupt when reached zero.
An add-on box could set this from the naval observatory. 370
(real-time) tod clock threads
http://www.garlic.com/~lynn/99.html#102 IBM 9020 computers used by FAA (was Re: EPO stories (was: HELP IT'S HOT!!!!!))
http://www.garlic.com/~lynn/2000.html#2 Computer of the century
http://www.garlic.com/~lynn/2000.html#4 Computer of the century
http://www.garlic.com/~lynn/2000f.html#57 X86 ultimate CISC? No. (was: Re: "all-out" vs less aggressive designs)
http://www.garlic.com/~lynn/2001f.html#52 any 70's era supercomputers that ran as slow as today's supercomputers?
http://www.garlic.com/~lynn/2001f.html#53 any 70's era supercomputers that ran as slow as today's supercomputers?
http://www.garlic.com/~lynn/2002.html#52 Microcode?

by comparison the 360/67 only had a 32bit decrementing timer that had
low-order bit position that was about 13+ microseconds (and would
interrupt at zero). cp/67 had to make due with process accounting and
time-slice control for interactive environment by doing a lot of
swapping and storing of that timer. it got easier for vm/370 on the
370 models with the real tod clock and the other time facilities.

real-time (hard processing time deadlines) wasn't really mainstream
mainframe ... there was the Federal Systems Division (FSD) special
computers done for nasa and other places that were real-time. then of
course there are the original modified 360/50s & 360/65s for the air
traffic control system. misc faa atc:
http://www.garlic.com/~lynn/2002b.html#6 Microcode?
http://www.garlic.com/~lynn/2002f.html#29 Computers in Science Fiction
http://www.garlic.com/~lynn/2002f.html#30 Computers in Science Fiction

then there was all the process controller stuff that was frequently
(real) real-time ... like 1800, system/7, & s/1 based stuff (of course
S/1 also saw a lot of use as communication controller in addition to
process controller). these were real-time in the sense that when
something happened in the real world, the computer had a (very) tight
realtime deadline to re-act/respond.

most of the mainframe stuff was real-time in the sense of handling
interactions with people (interactive as opposed to batch) ... but
also like ATM machines.

the one mainframe application (that I know of) that had real,
real-time constraints was the 1419 check sorter ... both MVT on 360
... and later VS1 on 370 had special I/O processing paths in order to
meet the elapsed time constraints of the 1419 check sorter.  misc
1419, check sorter:
http://www.garlic.com/~lynn/99.html#136a checks (was S/390 on PowerPC?)
http://www.garlic.com/~lynn/99.html#155 checks (was S/390 on PowerPC?)
http://www.garlic.com/~lynn/2002.html#18 Infiniband's impact was Re: Intel's 64-bit strategy

ACP (airline control program, currently called TPF) to maintain its
transaction rate ... did have stuff that killed transactions if they
exceeded a very small amount of processing. In the early '90s some of
the ACP/TPF airline systems had grown into pretty large clusters
handling several thousand transactions per second.

Some of the ATM & point-of-sale stuff have some interesting
characteristics (some also using TPF currently). In the '70s there was
a guy that wrote a virtual machine based ATM transaction processor for
VM/370 for one lf the LA area financial institutions ... he claimed
that he got higher ATM transaction throughput with vm/370 on a 370/158
(about 1mip processor) than TPF on 370/168 (about 3mip processor). The
difference was that he had a much more sophisticated arm queueing and
account record clustering strategies. recent acp/tpf thread:
http://www.garlic.com/~lynn/2002f.html#60 mainframes and mini-computers

A problem with ACP (& TPF) was that it was much closer to a monitor
than an operating system (with lots of sophisticated system
services). At least some of the large-scale airline res systems ... do
much of the prep & maint processing on auxiliary large scale MVS-based
database systems ... and then shutdown TPF periodically and rebuild
TPF infrastructure from the MVS hosted data.

As a independent consultant in the mid-90s we looked at porting routes
(res application accounting for something like 25 percent of total
activity) to a ha/cluster configuration. misc. routes
http://www.garlic.com/~lynn/96.html#31 Mainframes & Unix
http://www.garlic.com/~lynn/99.html#17 Old Computers
http://www.garlic.com/~lynn/99.html#100 Why won't the AS/400 die? Or, It's 1999 why do I have to learn how to use
http://www.garlic.com/~lynn/99.html#103 IBM 9020 computers used by FAA (was Re: EPO stories (was: HELP IT'S HOT!!!!!))
http://www.garlic.com/~lynn/99.html#136a checks (was S/390 on PowerPC?)
http://www.garlic.com/~lynn/99.html#153 Uptime (was Re: Q: S/390 on PowerPC?)
http://www.garlic.com/~lynn/2000.html#61 64 bit X86 ugliness (Re: Williamette trace cache (Re: First view of Willamette))
http://www.garlic.com/~lynn/2000e.html#19 Is Al Gore The Father of the Internet?^
http://www.garlic.com/~lynn/2000f.html#20 Competitors to SABRE?
http://www.garlic.com/~lynn/2001.html#26 Disk caching and file systems.  Disk history...people forget
http://www.garlic.com/~lynn/2001d.html#69 Block oriented I/O over IP
http://www.garlic.com/~lynn/2001g.html#45 Did AT&T offer Unix to Digital Equipment in the 70s?
http://www.garlic.com/~lynn/2001g.html#49 Did AT&T offer Unix to Digital Equipment in the 70s?
http://www.garlic.com/~lynn/2001j.html#17 I hate Compaq
http://www.garlic.com/~lynn/2001k.html#26 microsoft going poof [was: HP Compaq merger, here we go again.]
http://www.garlic.com/~lynn/2001n.html#0 TSS/360
http://www.garlic.com/~lynn/2001n.html#3 News IBM loses supercomputer crown

misc 1800, system/7, series/1
http://www.garlic.com/~lynn/94.html#8 scheduling & dynamic adaptive ... long posting warning
http://www.garlic.com/~lynn/99.html#63 System/1 ?
http://www.garlic.com/~lynn/99.html#64 Old naked woman ASCII art
http://www.garlic.com/~lynn/99.html#66 System/1 ?
http://www.garlic.com/~lynn/99.html#67 System/1 ?
http://www.garlic.com/~lynn/99.html#70 Series/1 as NCP (was: Re: System/1 ?)
http://www.garlic.com/~lynn/99.html#106 IBM Mainframe Model Numbers--then and now?
http://www.garlic.com/~lynn/99.html#239 IBM UC info
http://www.garlic.com/~lynn/2000b.html#66 oddly portable machines
http://www.garlic.com/~lynn/2000b.html#87 Motorola/Intel Wars
http://www.garlic.com/~lynn/2000c.html#35 What level of computer is needed for a computer to Love?
http://www.garlic.com/~lynn/2000c.html#43 Any Series/1 fans?
http://www.garlic.com/~lynn/2000c.html#51 WHAT IS A MAINFRAME???
http://www.garlic.com/~lynn/2000c.html#53 Any Series/1 fans?
http://www.garlic.com/~lynn/2001.html#4 Sv: First video terminal?
http://www.garlic.com/~lynn/2001.html#62 California DMV
http://www.garlic.com/~lynn/2001.html#65 California DMV
http://www.garlic.com/~lynn/2001.html#68 California DMV
http://www.garlic.com/~lynn/2001.html#72 California DMV
http://www.garlic.com/~lynn/2001b.html#75 Z/90, S/390, 370/ESA (slightly off topic)
http://www.garlic.com/~lynn/2001d.html#66 Pentium 4 Prefetch engine?
http://www.garlic.com/~lynn/2001d.html#69 Block oriented I/O over IP
http://www.garlic.com/~lynn/2001f.html#22 Early AIX including AIX/370
http://www.garlic.com/~lynn/2001f.html#30 IBM's "VM for the PC" c.1984??
http://www.garlic.com/~lynn/2001f.html#33 IBM's "VM for the PC" c.1984??
http://www.garlic.com/~lynn/2001f.html#42 Golden Era of Compilers
http://www.garlic.com/~lynn/2001f.html#44 Golden Era of Compilers
http://www.garlic.com/~lynn/2001i.html#21 3745 and SNI
http://www.garlic.com/~lynn/2001i.html#31 3745 and SNI
http://www.garlic.com/~lynn/2001k.html#8 Minimalist design (was Re: Parity - why even or odd)
http://www.garlic.com/~lynn/2001k.html#43 Why is UNIX semi-immune to viral infection?
http://www.garlic.com/~lynn/2001n.html#9 NCP
http://www.garlic.com/~lynn/2001n.html#23 Alpha vs. Itanic:  facts vs. FUD
http://www.garlic.com/~lynn/2001n.html#43 IBM 1800
http://www.garlic.com/~lynn/2001n.html#52 9-track tapes (by the armful)
http://www.garlic.com/~lynn/2001n.html#71 Q: Buffer overflow
http://www.garlic.com/~lynn/2002.html#7 The demise of compaq
http://www.garlic.com/~lynn/2002.html#20 Younger recruits versus experienced veterans  ( was Re: The demise  of compa
http://www.garlic.com/~lynn/2002.html#45 VM and/or Linux under OS/390?????
http://www.garlic.com/~lynn/2002b.html#56 Computer Naming Conventions
http://www.garlic.com/~lynn/2002c.html#18 Did Intel Bite Off More Than It Can Chew?
http://www.garlic.com/~lynn/2002c.html#42 Beginning of the end for SNA?
http://www.garlic.com/~lynn/2002d.html#16 Mainframers: Take back the light (spotlight, that is)

--
Anne & Lynn Wheeler   | lynn@garlic.com -  http://www.garlic.com/~lynn/

Why are Mainframe Computers really still in use at all?

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Why are Mainframe Computers really still in use at all?
Newsgroups: bit.listserv.ibm-main,alt.folklore.computers
Date: Mon, 29 Apr 2002 13:43:40 GMT

ss57@YAHOO.COM (Steve Smith) writes:

LexisNexis search engines and core data are on OS/390 mainframes.

i believe also dialog and some number of the large airline res systems
(although many for a long time had extended clusters using
HYPERChannel, including lexis/nexis and large tpf res systems)).

then there is possibly the largest & oldest, granddaddy of them all,
national library of medicine. NLM hit a huge bimodal problem early on
(70s) ... a search would return hundreds of thousands or more hits
... until somewhere out in the 5-7 term refinement, where the number
of hits dropped to zero.  Somewhere around 1980, a pc-based user
interface was developed to help manage search strategy ("grateful
med") and the default for NLM was to return the number of hits (as
opposed to the hits themselves).  Then it became something of a random
walk using grateful med ... to find a search combination that returned
less than hundreds or thousands but more than zero. as of 5-6 years
ago (last time I checked) ... the implementation was still late '60s
BDAM.

http://www.nlm.nih.gov/

--
Anne & Lynn Wheeler   | lynn@garlic.com -  http://www.garlic.com/~lynn/

markup vs wysiwyg (was: Re: learning how to use a computer)

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: markup vs wysiwyg (was: Re: learning how to use a computer)
Newsgroups: comp.sys.intel,comp.os.vms,comp.arch,alt.folklore.computers
Date: Mon, 29 Apr 2002 15:38:40 GMT

"Bob Knowles" writes:

And a bit more history: DEC (specifically CUP Engineering, now sadly
defunct) was part of the drive to get SGML adopted as a standard by ISO. At
what was thought to be the last hurdle, there was a delay (some of the
reviewers decided on a change that turned out to be more far-reaching than
they thought), and the schedule slipped by over a year.

i believe one of the players that pushed hard to turn GML into SGML
standard was the federal government & some guy that coined the term
COTS.

various gov. documents are really complex and they wanted some way of
transitioning to online. later when i was doing some due diligence on
tymshare for the m/d purchase ... one of the systems (besides
gnosis/keykos) was engelbart's augment ... which had some really
complex online gov. documents (i believe RFP responses) that had
really complex hyperlink structures (running on a tymshare tops-10
system).  there was also a cord keyboard that was in use with augment.

misc. augment posts:
http://www.garlic.com/~lynn/2000g.html#22 No more innovation?  Get serious
http://www.garlic.com/~lynn/2000g.html#26 Who Owns the HyperLink?
http://www.garlic.com/~lynn/2000g.html#31 stupid user stories

--
Anne & Lynn Wheeler   | lynn@garlic.com -  http://www.garlic.com/~lynn/

Black magic in POWER5

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Black magic in POWER5
Newsgroups: comp.arch
Date: Mon, 29 Apr 2002 16:10:03 GMT

Peter Boyle writes:

The link posted made reference to tcp/ip and MPI acceleration, so
if I were to make a crazy guess:

The memory controller goes on chip, and chips system bus interface has a
slave with bus qualifiers specifying PID and a virtual adressing mode. The
cpu could then translate the address (major faults and all), and you have
"acceleration" for dma from user space without the Myrinet style
pre-registration and locking of pages.

no idea, but one of the things from mainframe is all the access
register stuff (some 20 years old) and controlled ability to access
areas in multiple virtual address space areas ... allowing (pointer)
data access w/o having to do the buffer copies (or using locked/pinned
real address).

w/o the access rules, base 801/rios already has the makings if there
were operations that directly supported specification for
segment-id/segment-offset combo (i.e. PID might be needed for access
rules, but the pre-unix 801 already had segment-id/segment-offset for
unique virtual address).

there is all the old protocol engine stuff for tcp/ip (and xtp) for
scatter/gather of the pieces ... with trailer protocol for crc
calculated on the fly.

previous power5/power6 ref:
http://www.garlic.com/~lynn/2002f.html#57 IBM competes with Sun w/new Chips

some old 801 segment-id discussions:
http://www.garlic.com/~lynn/98.html#26 Merced & compilers (was Re: Effect of speed ... )
http://www.garlic.com/~lynn/2000.html#59 Multithreading underlies new development paradigm
http://www.garlic.com/~lynn/2000b.html#54 Multics dual-page-size scheme

misc. access register:
http://www.garlic.com/~lynn/98.html#36 What is MVS/ESA?
http://www.garlic.com/~lynn/2000d.html#28 RS/6000 vs. System/390 architecture?
http://www.garlic.com/~lynn/2001d.html#28 Very CISC Instuctions (Was: why the machine word size ...)
http://www.garlic.com/~lynn/2001d.html#30 Very CISC Instuctions (Was: why the machine word size ...)
http://www.garlic.com/~lynn/2001h.html#73 Most complex instructions
http://www.garlic.com/~lynn/2001k.html#16 Minimalist design (was Re: Parity - why even or odd)
http://www.garlic.com/~lynn/2002d.html#51 Hardest Mistake in Comp Arch to Fix

some old tcp/ip accelerator refs:
http://www.garlic.com/~lynn/99.html#0 Early tcp development?
http://www.garlic.com/~lynn/2000b.html#9 "Mainframe" Usage
http://www.garlic.com/~lynn/2001b.html#57 I am fed up!
http://www.garlic.com/~lynn/2001e.html#24 Pre ARPAnet email?
http://www.garlic.com/~lynn/2001h.html#22 Intel's new GBE card?
http://www.garlic.com/~lynn/2001k.html#62 SMP idea for the future

--
Anne & Lynn Wheeler   | lynn@garlic.com -  http://www.garlic.com/~lynn/

The HTML one-pixel dilation trick

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: The HTML one-pixel dilation trick
Newsgroups: alt.folklore.computers,comp.infosystems.www.authoring.html
Date: Mon, 29 Apr 2002 16:32:25 GMT

Brian Inglis writes:

Extracted from "The Mosaic Handbook", ORA, 1994.
Marc Andreessen left NCSA in Dec 93. NCSA agreed to have Spyglass
license and develop Mosaic in Aug 94. Sometime around there MA
and Jim Clark founded Mosaic Communications Corp.

other misc:
http://www.garlic.com/~lynn/aadsm5.htm#asrn2

--
Anne & Lynn Wheeler   | lynn@garlic.com -  http://www.garlic.com/~lynn/

income tax [was: Computers in Science Fiction]

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: income tax [was: Computers in Science Fiction]
Newsgroups: alt.folklore.computers
Date: Mon, 29 Apr 2002 22:39:08 GMT

William Hamblen writes:

Withholding first arrived with the Social Security tax in the '30s.
Income tax withholding arrived during WWII when the rates were raised
drastically.  Income tax day was changed from March 15 to April 15 at
the same time.

slightly more computer related is EFTPS (electronic funds tax payment
system) ... all corporations with greater than $200k/annum in
corporate tax + employee withholding have to use EFTPS (but anybody
can, there has also been pending regs to lower the threshold to
$50k/annum). Something like 1 million out of the 7 million
corporations in the US now use EFTPS. Also the treasury has been
quoted that something over 90 percent of the federal budget now flows
in via EFTPS. Some number of the states have also implemented similar
programs.

http://www.fms.treas.gov/eftps/index.html
http://www.fms.treas.gov/eftps/news.html

--
Anne & Lynn Wheeler   | lynn@garlic.com -  http://www.garlic.com/~lynn/

The HTML one-pixel dilation trick

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: The HTML one-pixel dilation trick
Newsgroups: alt.folklore.computers,comp.infosystems.www.authoring.html
Date: Tue, 30 Apr 2002 04:11:21 GMT

Arjun Ray writes:

Mosaic Communications Corp -> Netscape Communications Corp
and  Mosaic Netscape -> Netscape Navigator.

there was a company in the valley that had the rights to the word
netscape for some time ... they suggested/provided it to the new
company as alternative to mosaic.

--
Anne & Lynn Wheeler   | lynn@garlic.com -  http://www.garlic.com/~lynn/

IBM MIcrochannel??

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: IBM MIcrochannel??
Newsgroups: alt.folklore.computers
Date: Tue, 30 Apr 2002 15:46:34 GMT

maynard@thebrain.conmicro.cx (Jay Maynard) writes:

I have a couple of IBM MCA RS/6000s. They make great platforms for the P/390
personal mainframe system, much more so than the PCI versions that followed;
you can stuff an MCA P/390 board in any MCA box, and it'll work, while the
PCI version is much, much pickier.

the problem with MCA for RS/6000s wasn't the protocol itself or even
doubling the rate ... it was having a common bus hardware with the
ps/2 and suggestion that the rs/6000s should use ps/2 adapter cards
instead of designing their own. there was a joke if the rs/6000s had
to use the ps/2 scsi disk adapter card ... it would run as slow as
ps/2.

there were benchmarks of the ps/2 16mbit MCA t/r adapter card against
the PC/RT designed 4mbit ISA t/r adapter card. While 16mbit t/r might
have had higher aggregate thruput ... the cards were designed for
office desk top thruput. Any single 16mbit MCA t/r card had lower
sustained thruput than the PC/RT designed 4mbit ISA t/r card. Not so
bad for random office desk top use ... but serious problem for
engineering client/server ... especially where the "server" needed to
sustain the aggregate of all the individual clients.

--
Anne & Lynn Wheeler   | lynn@garlic.com -  http://www.garlic.com/~lynn/

"Soul of a New Machine" Computer?

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: "Soul of a New Machine" Computer?
Newsgroups: alt.folklore.computers
Date: Tue, 30 Apr 2002 15:31:55 GMT

Pete Fenelon writes:

It contributed significantly to DG revenue through the early 80s, though
largely in terms of keeping customer loyalty rather than being a "Vax
killer". DG later abandoned proprietary minis and got into Motorola
88k-based workstation/server systems running a variety of Unix
(DG/UX). The 88k was something of a turkey. DG almost died, but
resurrected itself as a company that builds Intel-based servers.
It was bought out by storage company EMC towards the back end of '99.

part of the emc purchase was that dg had gotten into disk arrays. dg
had built new campus with three new bldgs out on 495 ... which got
sold off (one of the bldgs bought by company that does the data
processing outsourcing for investment houses). DG, sequent, and convex
all built numa scaleup configurations with SCI interconnect (DG and
sequent with intel processors and i believe dolphin chips, convex with
hp processors and their own SCI design). dg got bought by emc (for its
clariion disk arrays?), sequent by ibm, and convex by hp.

rolm was using dg computers. we got to go in to look at the
development process. There was one stage where it was taking 24hrs
elapsed time just to load the new software into a switch for testing.
at one time the rolm campus across from great america was considered
the epitome of silicon valley. rolm was bought by ibm and then sold
off to siemens. siemens now has a couple other bldgs in the area
... including the complex at the first street & 101 for its infineon
spin-off.

somebody recently told a story about dg founder (de castro) and
getting his startup funding from the money people.

random dg refs:
http://computernewsdaily.com/247_090497_094212_26108.html
http://www.forbes.com/1997/10/14/feat.html
http://multics.acms.org.au/z0105.htm

misc. SCI refs:
http://www.garlic.com/~lynn/96.html#8 Why Do Mainframes Exist ???
http://www.garlic.com/~lynn/96.html#25 SGI O2 and Origin system announcements
http://www.garlic.com/~lynn/98.html#40 Comparison Cluster vs SMP?
http://www.garlic.com/~lynn/2001b.html#39 John Mashey's greatest hits
http://www.garlic.com/~lynn/2001b.html#85 what makes a cpu fast
http://www.garlic.com/~lynn/2001f.html#11 Climate, US, Japan & supers query
http://www.garlic.com/~lynn/2001j.html#12 OT - Internet Explorer V6.0
http://www.garlic.com/~lynn/2001j.html#17 I hate Compaq

--
Anne & Lynn Wheeler   | lynn@garlic.com -  http://www.garlic.com/~lynn/

"Soul of a New Machine" Computer?

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: "Soul of a New Machine" Computer?
Newsgroups: alt.folklore.computers
Date: Tue, 30 Apr 2002 15:38:57 GMT

cbh@ieya.co.REMOVE_THIS.uk (Chris Hedley) writes:

I didn't consider the 88k to be too bad... if anything, I thought
it was quite a decent processor.  Its main problem seemed to be
that Motorola lost interest when it started pursuing the Power
architecture.

or a lot of the 88k showed up in the power/pc. the rios/801/power was
a non-smp chip (no provisions for cache coherence). ok, i'm having a
senior moment ... what was the name of the joint group responsible for
601 ... our former boss headed it up (and he had come from motorola
before working on rios, later he was president of mips).

--
Anne & Lynn Wheeler   | lynn@garlic.com -  http://www.garlic.com/~lynn/

"Soul of a New Machine" Computer?

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: "Soul of a New Machine" Computer?
Newsgroups: alt.folklore.computers
Date: Tue, 30 Apr 2002 16:41:46 GMT

Anne & Lynn Wheeler writes:

a non-smp chip (no provisions for cache coherence). ok, i'm having a
senior moment ... what was the name of the joint group responsible for
601 ... our former boss headed it up (and he had come from motorola
before working on rios, later he was president of mips).

oops, finally clicked, somerset; misc. past threads.
ttp://www.garlic.com/~lynn/2000d.html#60 "all-out" vs less aggressive designs (was: Re: 36 to 32 bit transition)
http://www.garlic.com/~lynn/2001g.html#23 IA64 Rocks My World
http://www.garlic.com/~lynn/2001i.html#28 Proper ISA lifespan?
http://www.garlic.com/~lynn/2001j.html#37 Proper ISA lifespan?

--
Anne & Lynn Wheeler   | lynn@garlic.com -  http://www.garlic.com/~lynn/

Secure Device Drivers

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Secure Device Drivers
Newsgroups: comp.arch
Date: Tue, 30 Apr 2002 17:18:58 GMT

hack@watson.ibm.com (hack) writes:

Btw, we were in such a situation when we added FBA support: this differed
enough from (then) 3330/40/50 support that we had to rethink a number of
things.  But we had manuals, and they were not secret!  The work was also
completely below the file system (which just assumed a space of blocks of
a size reported at logical device connection time), so although it was not
all in ONE place, it was in a small number (3 or 4) of expected places.

When most people think of S/370 or its successors, they think of MVS and
its descendants, where (due to OS/360 heritage) many device properties
(in particular CKD-ness and track sizes) were visible all the way up into
application code, with assumptions spread all over the place.  That is
indeed a predicament that is difficult to outgrow -- but it is a property
of historical software conventions, not of the I/O interface.  Other S/390
operating systems such as VM, VSE, Linux or mine need not be burdened by it.

the problem with MVS supporting FBA was that it was really tied to CKD
& multi-track searching for both VTOC and PDS members. Even offering
them fully operational & tested support ... the STL group turned it
down, claimed the bill for just documentation and education in the
release cycle was on the order of $26m. I've claimed that the
life-time costs associated w/ECKD far surpases what it would have been
for FBA cut-over and its subsequent life-time costs.

It was easy for CP & CMS ... since both CP & CMS effectively treated
CKD devices as logical fixed block since the days of CP/40 in the
mid-60s.

past multi-track search threads
http://www.garlic.com/~lynn/93.html#29 Log Structured filesystems -- think twice
http://www.garlic.com/~lynn/94.html#35 mainframe CKD disks & PDS files (looong... warning)
http://www.garlic.com/~lynn/97.html#16 Why Mainframes?
http://www.garlic.com/~lynn/97.html#29 IA64 Self Virtualizable?
http://www.garlic.com/~lynn/99.html#75 Read if over 40 and have Mainframe  background
http://www.garlic.com/~lynn/2000.html#86 Ux's good points.
http://www.garlic.com/~lynn/2000f.html#18 OT?
http://www.garlic.com/~lynn/2000f.html#19 OT?
http://www.garlic.com/~lynn/2000f.html#42 IBM 3340 help
http://www.garlic.com/~lynn/2000g.html#51 > 512 byte disk blocks (was: 4M pages are a bad idea)
http://www.garlic.com/~lynn/2000g.html#52 > 512 byte disk blocks (was: 4M pages are a bad idea)
http://www.garlic.com/~lynn/2001.html#54 FBA History Question (was: RE: What's the meaning of track overfl ow?)
http://www.garlic.com/~lynn/2001b.html#82 Disks size growing while disk count shrinking = bad performance
http://www.garlic.com/~lynn/2001c.html#17 database (or b-tree) page sizes
http://www.garlic.com/~lynn/2001d.html#60 VTOC/VTOC INDEX/VVDS and performance (expansion of VTOC position)
http://www.garlic.com/~lynn/2001d.html#64 VTOC/VTOC INDEX/VVDS and performance (expansion of VTOC position)
http://www.garlic.com/~lynn/2001g.html#32 Did AT&T offer Unix to Digital Equipment in the 70s?
http://www.garlic.com/~lynn/2001l.html#40 MVS History (all parts)
http://www.garlic.com/~lynn/2002.html#5 index searching
http://www.garlic.com/~lynn/2002.html#6 index searching
http://www.garlic.com/~lynn/2002.html#10 index searching
http://www.garlic.com/~lynn/2002d.html#22 DASD response times
http://www.garlic.com/~lynn/2002f.html#8 Is AMD doing an Intel?

--
Anne & Lynn Wheeler   | lynn@garlic.com -  http://www.garlic.com/~lynn/

"Soul of a New Machine" Computer?

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: "Soul of a New Machine" Computer?
Newsgroups: alt.folklore.computers
Date: Wed, 01 May 2002 16:06:33 GMT

Charles Richmond writes:

Of course it looks "IBM-y"...it is based on the IBM Power
architecture of the RS-6000. It was jointly developed by
Apple, Monorola, and IBM, so no surprise there.

Someone must have thought that the PowerPC chips made good
embedded processors...they were used on the Mars Pathfinder
project...

somerset was the design team. it was headed up by guy that went on to
be president of mips (he had previously been at motorola before
joining the RIOS group for the power chips ... 32bit 810). the powerpc
workstation chips were 601, 603, 610, 615, 620, etc. Then there was
the rochester group doing what had been called the 630 for the
as/400. The 4xx chips were targeted at embedded market ... and
somewhere there was even talk of 2xx line of chips.

random past power/pc, somerset postings
http://www.garlic.com/~lynn/93.html#8 PowerPC Architecture (was: Re: PowerPC priced very low!)
http://www.garlic.com/~lynn/93.html#10 PowerPC Architecture (was: Re: PowerPC priced very low!)
http://www.garlic.com/~lynn/94.html#47 Rethinking Virtual Memory
http://www.garlic.com/~lynn/95.html#11 801 & power/pc
http://www.garlic.com/~lynn/98.html#25 Merced & compilers (was Re: Effect of speed ... )
http://www.garlic.com/~lynn/98.html#31 PowerPC MMU
http://www.garlic.com/~lynn/2000b.html#54 Multics dual-page-size scheme
http://www.garlic.com/~lynn/2000c.html#4 TF-1
http://www.garlic.com/~lynn/2000d.html#60 "all-out" vs less aggressive designs (was: Re: 36 to 32 bit transition)
http://www.garlic.com/~lynn/2001.html#38 Competitors to SABRE?
http://www.garlic.com/~lynn/2001.html#71 what is interrupt mask register?
http://www.garlic.com/~lynn/2001b.html#42 John Mashey's greatest hits
http://www.garlic.com/~lynn/2001g.html#13 Apple/PowerPC rumors
http://www.garlic.com/~lynn/2001g.html#23 IA64 Rocks My World
http://www.garlic.com/~lynn/2001i.html#24 Proper ISA lifespan?
http://www.garlic.com/~lynn/2001i.html#28 Proper ISA lifespan?
http://www.garlic.com/~lynn/2001j.html#36 Proper ISA lifespan?
http://www.garlic.com/~lynn/2001j.html#37 Proper ISA lifespan?
http://www.garlic.com/~lynn/2002c.html#19 Did Intel Bite Off More Than It Can Chew?
http://www.garlic.com/~lynn/2002c.html#38 Wang tower minicomputer
http://www.garlic.com/~lynn/2002g.html#11 "Soul of a New Machine" Computer?
http://www.garlic.com/~lynn/2002g.html#12 "Soul of a New Machine" Computer?

--
Anne & Lynn Wheeler   | lynn@garlic.com -  http://www.garlic.com/~lynn/

Security Issues of using Internet Banking

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Security Issues of using Internet Banking
Newsgroups: alt.computer.security,alt.folklore.computers
Date: Wed, 01 May 2002 16:15:07 GMT

mrr@acer.reistad.priv.no (Morten Reistad) writes:

It is user-friendly and VERY lightweight.(it works OK over a 9600 bps mobile
phone connection; with image loading turned off).

one of the requirements given the X9A10 working group for the x9.59
payment standard was that it work for all electronic payments in all
environments.

one of the environmental issues are point-of-sale terminals. a large
percentage of POS terminals are kind-of repackaged PC/XTs with 2400
baud modems. One of the recent issues that came up looking at
upgrading POS terminals to higher speeds ... was that just the
protocol initialization hand-shaking chatter for higher speed modems
takes longer than total elapsed time for existing calls.

misc. x9.59 refs:
http://www.garlic.com/~lynn/x959.html#x959

--
Anne & Lynn Wheeler   | lynn@garlic.com -  http://www.garlic.com/~lynn/

Why are Mainframe Computers really still in use at all?

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Why are Mainframe Computers really still in use at all?
Newsgroups: bit.listserv.ibm-main,alt.folklore.computers
Date: Wed, 01 May 2002 16:19:46 GMT

eugene@cse.ucsc.edu (Eugene Miya) writes:

Good question.
Ask the IRS.

i think that IRS has had more re-engineering/moderniation projects
over the last 15 years than even FAA ATC. The one at the census was
somewhat more succesful for the year 2000 census ... but they have 10
years between events to get ready.

--
Anne & Lynn Wheeler   | lynn@garlic.com -  http://www.garlic.com/~lynn/

Black magic in POWER5

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Black magic in POWER5
Newsgroups: comp.arch
Date: Wed, 01 May 2002 16:48:34 GMT

Peter Boyle writes:

i.  I/O can assert virtual addresses (whether or not locked) and have
translation carried out, so no pregistration of pages and/or g-s
tables. I'm ashamed to admit I was entirely unaware of the
mainframe segment translations the Wheelers posted, they don't really
work in a typical contemporary VM system.

hardware access registers was ability to have multiple virtual address
spaces defined for a process ... with rules about accessing data in
other than the primary virtual address space. part of this was
targeted at semi-priviledged processes providing system services
... but running in their own address space ... and being able to
directly access parameters and other passed data. Furthermore, it is
possible to do a "call" & context switch to semi-priviledge processes
in other address spaces w/o having to take an interrupt into the
kernel and execute kernel instructions. The makings are there for
demon like processes & other services in their own, unique address
spaces that are directly calleable from an application program. The
mechanisms are then there for having hardware support for a collection
of data ... where different pieces of the collection resided in
different virtual address spaces. architecture for access registers
and multiple address space technology date back to the late '70s. The
earliest implementation was "dual-address" space support on the 3033
in the late '70s.

somewhat in parallel with the multiple address space architecture
support in the mainframe was the 801 work dating from the same era.
This showed up in the ROMP chip ... used in PC/RTs. ROMP, instead of
having a virtual address space register ... had 16 segment registers
in a 32bit virtual address space. The top four bits of the 32bit
virtual address selected/indexed a segment register. A ROMP segment
register contained a 12bit segment-ID. The hardware table look aside,
operated with a combo of 12bit segment-id/16bit page number. The ROMP
documentation talked about 40bit virtual addressing ... this is
because the hardware had 28-bit virtual address displacements within a
segment plus the 12bit segment-id. A ROMP processor, in theory could
have up to 4096 simultaneously actively defined virtual memory
segments. If this was mapped into 32-bit unique virtual address spaces
with 16 privately defined 256mbyte segments per 32-bit virtual address
space ... it allowed for 256 simultaneously actively defined virtual
address spaces supported in hardware. Note however, that the 801/ROMP
architecture/implementation wasn't originally targeted at a UNIX
virtual address space paradigm.

RIOS/POWER for the RS/6000 extended the 12bit segment-id found in ROMP
to 24bits ... and therefor the RIOS/POWER documentation mentioning
52-bit virtual addressing supported by the hardware.

Part of the original issue involving things like buffer copies and
different domains of control ... would involve the hardware being able
to simultaneously access different pieces of data appearing in
different address space ... and have an address specification
mechanism that included multiple address spaces. The mainframe
hardware access registers provide such a mechanism (original
implementation in the late '70s with the 3033 dual-address space
mode). The ROMP/RIOS also support hardware unique addressing across
multiple virtual address spaces with "segment-ids".

A case could even be made that both the ROMP/RIOS and the access
register solutions originally evolved in nearly the same time & place
... even with some of the same people overlap.

original post:
http://www.garlic.com/~lynn/2002g.html#5

--
Anne & Lynn Wheeler   | lynn@garlic.com -  http://www.garlic.com/~lynn/

Black magic in POWER5

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Black magic in POWER5
Newsgroups: comp.arch
Date: Wed, 01 May 2002 17:01:41 GMT

Anne & Lynn Wheeler writes:

A case could even be made that both the ROMP/RIOS and the access
register solutions originally evolved in nearly the same time & place
... even with some of the same people overlap.

simple example:
http://www.hpl.hp.com/news/2001/apr-jun/2worley.html
http://web.archive.org/web/20021226021851/www.hpl.hp.com/news/2001/apr-jun/2worley.html

see dual-address space reference in the above.

--
Anne & Lynn Wheeler   | lynn@garlic.com -  http://www.garlic.com/~lynn/

Why did OSI fail compared with TCP-IP?

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Why did OSI fail compared with TCP-IP?
Newsgroups: comp.arch,comp.protocols.iso,comp.protocols.tcp-ip,comp.lang.c++,alt.folklore.computers
Date: Wed, 01 May 2002 20:19:17 GMT

"Rudvar Alswill" writes:

OSI was started in the seventies to find a model and develop protocols
for computer communications. At a cost of almost 1 Billion dollars and
twenty years, it failed completely.
It was strongly supported by Telecom groups in Europe. Yet they were
totally discredited by producing virtually nothing in that time and for
that amount of money.
How did so many highly technical people waste so much time and money and
fail in such a undignified way.

1) OSI was strongly oriented towards the telco copper wire
point-to-point problems of the 70s & earlier, like high error rates
and little or no FEC technology. it was also pre-LAN

2) TCP/IP was a working implementation ... where OSI tended to be a
lot of specification independent of practical implementation ... IETF
at least required two operational, interoperable implementations prior
to advancement to RFC.

3) TCP/IP underwent major evolution with the 1/1/83 cut-over from the
strongly host-to-host orientation into "internet" (aka IP part of
tcp/ip).

in the late '80s at one point in X3S3.3 (ansi standards for equivalent
to OSI level 3 & 4) ... work on high speed protocol (HSP) effort was
fealt to be very dubious because progression to ISO level supposedly
required conformance with seperation of level 3 & level 4
operations. HSP would have collapsed portions of level 3 & 4 into
single level. The ISO & ANSI OSI-related groups were already quite
skizo over this requirement since IEEE 803 had already collapsed OSI
level 1, level 2, and parts of level 3 into single layer (and there
was no obvious easy way of declaring LANs invalid and having them all
destroyed). HSP would have filled between the level 4 interface to
IEEE 803 (aka all of OSI level 4 and all of OSI level 3 not already
occupied by IEEE 803).

in that sense the organizational activities around the "pure,
original, OSI architecture" wasn't very agile at adapting to changing
technology. There were lots of organizational mandates about meeting
pure architecture specification ... but not a whole lot of attention
to practicallities of the real market and changing technology.

random past postings on this subject:
http://www.garlic.com/~lynn/99.html#0 Early tcp development?
http://www.garlic.com/~lynn/99.html#114 What is the use of OSI Reference Model?
http://www.garlic.com/~lynn/99.html#115 What is the use of OSI Reference Model?
http://www.garlic.com/~lynn/2000b.html#0 "Mainframe" Usage
http://www.garlic.com/~lynn/2000b.html#1 "Mainframe" Usage
http://www.garlic.com/~lynn/2000b.html#4 "Mainframe" Usage
http://www.garlic.com/~lynn/2000b.html#5 "Mainframe" Usage
http://www.garlic.com/~lynn/2000b.html#8 "Mainframe" Usage
http://www.garlic.com/~lynn/2000b.html#9 "Mainframe" Usage
http://www.garlic.com/~lynn/2000b.html#10 "Mainframe" Usage
http://www.garlic.com/~lynn/2000b.html#59 7 layers to a program
http://www.garlic.com/~lynn/2000b.html#79 "Database" term ok for plain files?
http://www.garlic.com/~lynn/2000d.html#63 Is Al Gore The Father of the Internet?
http://www.garlic.com/~lynn/2000d.html#70 When the Internet went private
http://www.garlic.com/~lynn/2000d.html#72 When the Internet went private
http://www.garlic.com/~lynn/2000e.html#19 Is Al Gore The Father of the Internet?^
http://www.garlic.com/~lynn/2001b.html#57 I am fed up!
http://www.garlic.com/~lynn/2001e.html#16 Pre ARPAnet email?
http://www.garlic.com/~lynn/2001e.html#17 Pre ARPAnet email?
http://www.garlic.com/~lynn/2001e.html#23 Pre ARPAnet email?
http://www.garlic.com/~lynn/2001e.html#24 Pre ARPAnet email?
http://www.garlic.com/~lynn/2001e.html#25 Pre ARPAnet email?
http://www.garlic.com/~lynn/2001e.html#32 Blame it all on Microsoft
http://www.garlic.com/~lynn/2001e.html#34 Blame it all on Microsoft
http://www.garlic.com/~lynn/2001i.html#5 YKYGOW...
http://www.garlic.com/~lynn/2001i.html#6 YKYGOW...
http://www.garlic.com/~lynn/2001j.html#4 I hate Compaq
http://www.garlic.com/~lynn/2001j.html#20 OT - Internet Explorer V6.0
http://www.garlic.com/~lynn/2001k.html#62 SMP idea for the future
http://www.garlic.com/~lynn/2001k.html#71 Encryption + Error Correction
http://www.garlic.com/~lynn/2001m.html#15 departmental servers
http://www.garlic.com/~lynn/2001n.html#15 Replace SNA communication to host with something else
http://www.garlic.com/~lynn/2001n.html#27 Unpacking my 15-year old office boxes generates memory refreshes
http://www.garlic.com/~lynn/2002e.html#53 Mainframers: Take back the light (spotlight, that is)
http://www.garlic.com/~lynn/2002e.html#61 Computers in Science Fiction

--
Anne & Lynn Wheeler   | lynn@garlic.com -  http://www.garlic.com/~lynn/

Security Issues of using Internet Banking

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Security Issues of using Internet Banking
Newsgroups: alt.computer.security,alt.folklore.computers
Date: Wed, 01 May 2002 21:21:01 GMT

"Charlie Gibbs" writes:

Right now I'd settle for having an ATM not lie to me.  That's what
I call it when it says my account balance is $1200, and then says
"your request exceeds your balance" when I try to withdraw $40.
(There was a hold on my account - which is another topic in itself -
but the machine couldn't be bothered telling me about that.)

getting a failure indication thru the debit network from the bank-end
bank processor out to the ATM machine is major undertaking
... frequently they map several decline/error indications into a
standard message.

for the NACHA aads trials (specifically "NACHA AADS RFI" & "NACHA AADS
results references"):
http://www.garlic.com/~lynn/x959.html#aads

it was a major under-taking to get the backend processor to return
"incorrect digital signature" in place of "incorrect PIN". I've seen
better worded message that says "your request exceeds your available
balance" ... i.e. "available" taken to be a whole set of conditions.
Of course somebody might have been trying to cut the number of
characters displayed. Also this can be distinct indicator from "your
request exceeds your limit" ... although both might map to same
condition ... retrying the operation with a lower value.

--
Anne & Lynn Wheeler   | lynn@garlic.com -  http://www.garlic.com/~lynn/

Why did OSI fail compared with TCP-IP?

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Why did OSI fail compared with TCP-IP?
Newsgroups: comp.arch,comp.protocols.iso,comp.protocols.tcp-ip,comp.lang.c++,alt.folklore.computers
Date: Thu, 02 May 2002 03:40:41 GMT

"David H. Lipman" <DLipman~nospam~@Verizon.Net> writes:

Do you remember GOSIP ?

Nobody wanted to "reinvent the wheel" after TCP/IP became so widely
proliferated.
{ GOSIP - Gov't. Open Systems Internet Protocol }

post containing some gosip extracts and document refs:
http://www.garlic.com/~lynn/2000d.html#70 When the Internet went private

a couple misc. files tripped across:
gosip-v2.txt ... also gosip-v2.ps ... 10/90
gosip-order-info.txt                   9/91
vendors-guide.doc                      8/90

RFCs mentioning GOSIP ... go to
http://www.garlic.com/~lynn/rfcietff.htm

& click on Term (term->RFC#)
then click on "GOSIP" in the Acronym fastpath
which gives:

  Government OSI Profile (GOSIP)
  see also Open Systems Interconnection
  2441 1632 1629 1237 1169 1039

clicking on the actual RFC number, will bring up the RFC summary in
the bottom frame. clicking on the ".txt=nnnnn" field in the summary will
retrieve that specific RFC. misc. summaries from above:

1039
DoD statement on Open Systems Interconnection protocols, Latham D.,
1988/01/01 (3pp) (.txt=6024) (Obsoletes 945)

1169
Explaining the Role of GOSIP, Cerf V., Mills K., 1990/08/09 (15pp)
(.txt=29413)

some of the previously refs mentioning gosip:
http://www.garlic.com/~lynn/99.html#114 What is the use of OSI Reference Model?
http://www.garlic.com/~lynn/99.html#115 What is the use of OSI Reference Model?
http://www.garlic.com/~lynn/2000b.html#0 "Mainframe" Usage
http://www.garlic.com/~lynn/2000b.html#59 7 layers to a program
http://www.garlic.com/~lynn/2000b.html#79 "Database" term ok for plain files?
http://www.garlic.com/~lynn/2000d.html#16 The author Ronda Hauben fights for our freedom.
http://www.garlic.com/~lynn/2000d.html#43 Al Gore: Inventing the Internet...
http://www.garlic.com/~lynn/2000d.html#63 Is Al Gore The Father of the Internet?
http://www.garlic.com/~lynn/2001e.html#17 Pre ARPAnet email?
http://www.garlic.com/~lynn/2001e.html#32 Blame it all on Microsoft
http://www.garlic.com/~lynn/2001i.html#5 YKYGOW...
http://www.garlic.com/~lynn/2001i.html#6 YKYGOW...

--
Anne & Lynn Wheeler   | lynn@garlic.com -  http://www.garlic.com/~lynn/

Why did OSI fail compared with TCP-IP?

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Why did OSI fail compared with TCP-IP?
Newsgroups: comp.arch,comp.protocols.iso,comp.protocols.tcp-ip,comp.lang.c++,alt.folklore.computers
Date: Thu, 02 May 2002 15:13:15 GMT

Jason Ozolins writes:

And don't forget, what was implemented tended to perform like slugs on
drugs.  In 1993 I watched a file transfer over Ethernet between a
Windows 3.1 box running ICL's beautiful office automation suite and
an ICL SPARC SVR4 box running OSI protocols go about the same speed as
my 9600 baud modem at home could manage.  At the time, my thought was
"if this is the future, then the future will be very very slow".

I believe that some organization in europe implemented the full
7-level OSI stack and when the thruput numbers were presented somebody
from one of the the ISO group said that OSI was supposed to be a
abstract model ... not a real specification.

--
Anne & Lynn Wheeler   | lynn@garlic.com -  http://www.garlic.com/~lynn/

Why are Mainframe Computers really still in use at all?

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Why are Mainframe Computers really still in use at all?
Newsgroups: alt.folklore.computers
Date: Thu, 02 May 2002 15:04:38 GMT

ab528@FreeNet.Carleton.CA (Heinz W. Wiggeshoff) writes:

What consulting outfit got their claws into the IRS?  I'm sure that they
got desperate enough at some point to farm out the smart work.

the most recent modernization round I believe has been CSC. however, i
believe there is lots of talk of another round involving joint bid
from ???  (there was some bid press release ... i think it was joint
bid with 3-4 companies with one being IBM). I believe I've heard some
comment that there is some amount of 360 assembler still in use that
has to be analyzed as part of any modernization effort.

I don't think it is an issue of desperation leading to farming out the
work.  I think it may be long term staffing levels ... for a
(supposedly) short term project of any magnitude do you hire a bunch
of civil serpents who you then have to fire when the project is done?
Furthermore, the skill mix & staffing levels for such a modernization
project is pretty different than the day-to-day operational & support
staff.

--
Anne & Lynn Wheeler   | lynn@garlic.com -  http://www.garlic.com/~lynn/

Why did OSI fail compared with TCP-IP?

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Why did OSI fail compared with TCP-IP?
Newsgroups: alt.folklore.computers
Date: Thu, 02 May 2002 15:18:07 GMT

"George R. Gonzalez" writes:

One personal experience-- we were trying to decide around 1989 whether to
go with X.500 for directory services (on a 640K DOS pc client no less).

All we really needed was a souped up "finger" daemon.  Easily understood
and implemented.

I downloaded the X.500 client and server-- memory fails me but it was
something like 33 megabytes of tar file, 1,300 files.

I remember being at a ACM conference around 1990 (sigmod?) ... and
somebody there characterizing x.500 as a bunch of networking types
re-inventing 1960s database technology.

--
Anne & Lynn Wheeler   | lynn@garlic.com -  http://www.garlic.com/~lynn/

Why are Mainframe Computers really still in use at all?

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Why are Mainframe Computers really still in use at all?
Newsgroups: alt.folklore.computers
Date: Thu, 02 May 2002 15:20:35 GMT

Anne & Lynn Wheeler writes:

I don't think it is an issue of desperation leading to farming out the
work.  I think it may be long term staffing levels ... for a
(supposedly) short term project of any magnitude do you hire a bunch
of civil serpents who you then have to fire when the project is done?
Furthermore, the skill mix & staffing levels for such a modernization
project is pretty different than the day-to-day operational & support
staff.

also can you imagine trying to get a shole bunch of short-term
supergrades thru congressional approval process?

--
Anne & Lynn Wheeler   | lynn@garlic.com -  http://www.garlic.com/~lynn/

Why did OSI fail compared with TCP-IP?

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Why did OSI fail compared with TCP-IP?
Newsgroups: comp.arch
Date: Thu, 02 May 2002 15:30:38 GMT

Chris Quayle writes:

It's easy now to have 20-20 vision now in hindsight, but networking at
the time was very much in it's infancy. Early computer networks were
based on a telecoms model because that was the working global comms
network of the time. The state machine design model came directly from
hardware design techniques. Some of the early protocols (X25 frame etc)
were implemented at hardware level because the computing power to do
this in software did not exist at a reasonable cost. Even now, state
machine design techniques are an excellent way to visualise and
implement complex protocols. The difference being that now many more of
the layers are implemented in software, rather than hardware.

It's surprising how the layered model and state machine design
techniques have persisted right up to the present time, so perhaps some
of those original ideas were more on target than is given credit for.

however, IETF adopted to the real-world ... ISO effectively kept
saying that IEEE 803 LANs violated the OSI model and shouldn't be
allowed to exist. we ran into that in x3s3.3 on HSP that was trying to
doing something that went directly from the top of level4 directly to
mid-level3 ... aka top of IEEE803 LANs and were told it would never
receive ISO standardization because it violated the OSI model.

--
Anne & Lynn Wheeler   | lynn@garlic.com -  http://www.garlic.com/~lynn/

Security Issues of using Internet Banking

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Security Issues of using Internet Banking
Newsgroups: alt.computer.security,alt.folklore.computers
Date: Thu, 02 May 2002 17:11:58 GMT

Brian Inglis writes:

If you can think up the test for the condition that leads to this
abrupt exit, the least you can do is tell the user or support
what the condition was that caused the exit, and provide a unique
error reference that leads back to the test in the code, even if
you don't know what caused it and what it means.
That's what C's __FILE__, __LINE__, and now __func__, are for,
together with a nice enum err { ERR_ }, to guide support right
to the defensive code, in case the error text is not sufficiently
descriptive.

ibm operating systems and products started that in the 60s, things of
the form: module, error code, severity level and verbal description.
for interactive like CMS ... it was possible to set option where
severity level could be filtered or message could be just the code,
just the verbage or both. then there were big manuals that listed each
along with paragraph or two more detailed description. something
similar was done for kernel panics.

I wrote a post-mortem kernel dump analyzer (in REXX, then called just
REX) that had softcopy of the messages&codes manual and would
automatically reference the appropriate detailed description.  I also
started development of a library of scripts that could do certain sets
of specific analysis based on failure code.

misc. dumprx refs:
http://www.garlic.com/~lynn/94.html#11 REXX
http://www.garlic.com/~lynn/2000b.html#32 20th March 2000
http://www.garlic.com/~lynn/2000b.html#33 20th March 2000
http://www.garlic.com/~lynn/2001c.html#0 Z/90, S/390, 370/ESA (slightly off topic)

--
Anne & Lynn Wheeler   | lynn@garlic.com -  http://www.garlic.com/~lynn/

Why did OSI fail compared with TCP-IP?

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Why did OSI fail compared with TCP-IP?
Newsgroups: comp.arch,comp.protocols.iso,comp.protocols.tcp-ip,comp.lang.c++,alt.folklore.computers
Date: Thu, 02 May 2002 17:31:14 GMT

Chris Morgan writes:

The obvious concrete example being LDAP, which is a subset of
X.something (500?)

as mentioned in another posting somebody at an ACM conference circa
1990 (SIGMOD?) ... x.500 was a bunch of networking types re-inventing
1960s database technology.

LDAP is much more of a networking protocol that accesses database
backends (you don't see many LDAPs using other than pre-existing DBMS
technology). It is also heavily influenced by IETF process ... aka

http://www.garlic.com/~lynn/rfcietff.htm

select Term (term->RFC#) and then select "LDAP" from the Acronym fastpath,
i.e.:

ightweight directory access protocol (LDAP ) (LDAPv2) (LDAPv3 )
see also ITU directory service protocol , directory
3112 3088 3062 3060 3045 2927 2926 2891 2849 2830 2829 2820 2798 2739
2714 2713 2696 2657 2649 2596 2589 2587 2559 2307 2256 2255 2254 2253
2252 2251 2247 2164 1960 1959 1823 1798 1778 1777 1558 1487 1249

it is then possible to select on any of the RFC numbers to get a summary
of the specific RFC. In the RFC summary entry it is possible to select
on the ".txt=nnnn" field to retrieve the actual RFC ... aka

3088 E
OpenLDAP Root Service An experimental LDAP referral service, Zeilenga
K., 2001/04/16 (11pp) (.txt=19471)

--
Anne & Lynn Wheeler   | lynn@garlic.com -  http://www.garlic.com/~lynn/

Why did OSI fail compared with TCP-IP?

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Why did OSI fail compared with TCP-IP?
Newsgroups: comp.arch,comp.protocols.iso,comp.protocols.tcp-ip,comp.lang.c++,alt.folklore.computer
Date: Thu, 02 May 2002 17:45:43 GMT

Tony Finch writes:

And SNMP uses ASN.1 and SSL certificates are X.509.

SSL certificates are sort-of X.509 ... and X.509 certificates are
specified as ASN.1 encoded, aka ASN.1 encoding is used for
transmission purposes. Once transmitted, to use the information, there
has to be ASN.1 decoding of the information. There is some move for
XML encoding of various things in place of ASN.1 encoding.

There are huge number of things that are ISO standards ... other than
OSI networking standards. Many such ISO standards are widely deployed
and succesful. Just because something is an ISO standard doesn't
automagically make it a failure and just because something is an IETF
standard doesn't automagically make it a success. However, IETF has
had something of a track record of requiring actual operational
implementations before moving things along the standards track, aka it
doesn't mean that the people in IETF are either smarter or dumber
... but there is something of a sanity check having real live
implementations.

A world-wide deployed network message protocol that is an ISO standard
is 8583 (neither TCP/IP nor OSI) ... all those ATM and point-of-sale
(debit & credit) boxes that you find all over the world (as well as
the backend bank-to-bank). Note, however it doesn't use ASN.1
encoding.

--
Anne & Lynn Wheeler   | lynn@garlic.com -  http://www.garlic.com/~lynn/

Why did OSI fail compared with TCP-IP?

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Why did OSI fail compared with TCP-IP?
Newsgroups: comp.arch,comp.protocols.iso,comp.protocols.tcp-ip,comp.lang.c++
Date: Thu, 02 May 2002 21:51:46 GMT

Marco S Hyman writes:

>{ GOSIP - Gov't. Open Systems Internet Protocol }

Not "protocol", "profile".   It was a functional profile of which of the
(some times incompatible) parts of OSI should be used for systems sold to
the US government.  Exanple: OSI had two different (and incompatible)
network protocols, CONS and CLNS.  GOSIP specified which one a conforming
implemention must use.

// marc

and not internet ... but interconnection ... aka

U. S. Government Open Systems Interconnection Profile (GOSIP)
VERSION 2.0
October 1990

the whole thing about internet (aka IP) was the great 1/1/83
switch-over for arpanet ... prior to that it looked more like
traditional (homogeneous protocol) network. the 1/1/83 switch-over
introduced the concept of internet and gateways (between networks).

one of my claims regarding the internal corporate network being larger
than all of arpanet/internet from the beginning until some time around
'85 was that effectively the internal corporate network essentially
supported gateway function in every node from the beginning.

misc. internet posts:
http://www.garlic.com/~lynn/internet.htm
http://www.garlic.com/~lynn/subtopic.html#internet
http://www.garlic.com/~lynn/subindex.html#network
http://www.garlic.com/~lynn/subindx2.html#network

misc. osi, hsp, iso posts:
http://www.garlic.com/~lynn/subtopic.html#xtphsp

--
Anne & Lynn Wheeler   | lynn@garlic.com -  http://www.garlic.com/~lynn/

Why did OSI fail compared with TCP-IP?

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Why did OSI fail compared with TCP-IP?
Newsgroups: comp.arch,comp.protocols.iso,comp.protocols.tcp-ip
Date: Fri, 03 May 2002 13:20:46 GMT

"Ketil Malde" <ketil+@ii.uib.no> writes:

You don't have to look to Adams; In "The fifth discipline", Peter
Senge considers a committee to effectively work at the lowest IQ of
the members, IIRC.  Good book, by the way.

different committees operate at different levels

sum(IQ0, IQ1, ..., IQn)
sum(IQ0, IQ1, ..., IQn)/n
min(IQ0, IQ1, ..., IQn)
min(IQ0, IQ1, ..., IQn)/n

i.e. a comittee can be less effective than any single member of the
comittee alone ... with large comittees tending to zero effectiveness.

--
Anne & Lynn Wheeler   | lynn@garlic.com -  http://www.garlic.com/~lynn/

Secure Device Drivers

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Secure Device Drivers
Newsgroups: comp.arch
Date: Fri, 03 May 2002 13:47:40 GMT

Jan C. Vorbrüggen writes:

I have the distinct impression OS implementors disagree with you. For
instance, VMS only fairly late offered this mechanism (called FAST_IO, IIRC
- there is also FAST_PATH, which AFAIK is more to do with making sure an
I/Os path through the OS is as short as possible) and it requires slightly
modified code on the side of the application. There are some restrictions
(I think buffers have to be page-aligned, for instance), and it is clearly
only worthwhile if you are going to do a lot of I/O. It does offer
substantially improved performance at some initial cost, both in programming
and at run time.

when i worked on redoing the i/o supervisor and error recovery to make
I/O bullet proof for the disk engineering lab (i.e. at the time, MVS
had something like 15 minute MTBF operating with a single test cell):
http://www.garlic.com/~lynn/subtopic.html#disk

the general path also got a lot faster ... in part because a lot of
the code was somewhat spaghetti ... having been the result of
something like 10+ year growth, modification, enhancements, and fixes
(as well as having numerous anomolous conditions).

but I also did some amount of fastpath for the optimized case
(somewhat out of habit having started doing fastpath something like 15
years earlier).

I also redid the path finding for i/o and path load balancing to
improve its optimization and pathlenth. and then I had to redo it
again. the emerging 3880 disk controller turned out to have
significant alternate path penalty overhead. If the 3880 had to change
channel/path between I/O operations there was significant internal
latency and overhead (not only latency to initiate operations but also
busy for servicing any other operations during the period). At low to
moderate loading levels ... multiple path overhead in the 3880 tended
to offset any increase in thruput because of additional parallel
transfer paths.  It wasn't until really high transfer utilization
(optimally few very large transfers, minimizing 3880 control op
involvement) that benefit of parallel transfers started to offset the
internal 3880 controller overheads.  As a result, parallel path
scheduling wasn't just a simple matter of straight-forward load
balancing across the available paths. The scenario at the time was
that there could be 16-32 3mbyte/sec disks connected to a 3880
controller with up to four possible 3mbyte/sec parallel transfer paths
(an alternative configuration was the same set of disks could be
connected to a pair of 3880 controllers, each with up to four
transfers paths, allowing up to eight parallel transfer paths to the
same set of disks, clustering of up to eight different processors,
each with a single path ... or some combination of processor
clustering and parallel paths).

The other issue at the time was trying to get MVS support for FBA
which significantly improved thruput in nearly all common
environments.

misc. previous fastpath refs:
http://www.garlic.com/~lynn/93.html#1 360/67, was Re: IBM's Project F/S ?
http://www.garlic.com/~lynn/94.html#2 Schedulers
http://www.garlic.com/~lynn/94.html#54 How Do the Old Mainframes
http://www.garlic.com/~lynn/97.html#22 Pre S/360 IBM Operating Systems?
http://www.garlic.com/~lynn/2002.html#14 index searching
http://www.garlic.com/~lynn/2002c.html#45 cp/67 addenda (cross-post warning)
http://www.garlic.com/~lynn/2002e.html#52 PKI and Relying Parties
http://www.garlic.com/~lynn/2002e.html#60 Browser Security
http://www.garlic.com/~lynn/2002f.html#57 IBM competes with Sun w/new Chips

recent FBA & CKD thruput postings:
http://www.garlic.com/~lynn/2002g.html#13 Secure Device Drivers

--
Anne & Lynn Wheeler   | lynn@garlic.com -  http://www.garlic.com/~lynn/

ESCON Distance Limitations - Why ?

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: ESCON Distance Limitations - Why ?
Newsgroups: comp.arch.storage
Date: Fri, 03 May 2002 14:50:21 GMT

"Yves Amblard" writes:

Can anyone explain the origin of the ESCON's distance limitations ?
If found a lot of documents related to ESCON but none is giving a
mathematical explanation.
I know it's due to the protocol and signal propagation but was unable to
find numbers or a mathematical explanation for it.
I found a lot of documents on the web describing the distance limitation
issue but no real explanation.

ESCON is a half-duplex, syncronized at the multi-byte level (not at
the per transfer or per operation level). especially with disks &
tapes there are various operations that presume continuous (or at
least within specific latency) operation once started.

It inherited the design from data-streaming half-duplex channel cables
(originally increasing the per byte syncronized transfer distance from
200' to 8-byte syncronized transfer distance limitation of
400'). Don't think of ESCON like dual-simplex FCS ... and operation
latencies ... think of it as data-streaming half-duplex bus&tag copper
that just happens to be using fiber.

try search engines with escon, data-streaming, distance; example
http://www-1.ibm.com/servers/eserver/zseries/library/specsheets/pdf/optica.pdf

or "escon extended distance"

for exact formulae, might try library for something like ibm r&d
journal looking for an article when data-streaming &/or escon were
originally announced.
http://www.research.ibm.com/journal/
http://www.research.ibm.com/journal/rd37-6.html

--
Anne & Lynn Wheeler   | lynn@garlic.com -  http://www.garlic.com/~lynn/

Why did OSI fail compared with TCP-IP?

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Why did OSI fail compared with TCP-IP?
Newsgroups: comp.arch,comp.protocols.iso,comp.protocols.tcp-ip
Date: Fri, 03 May 2002 17:54:55 GMT

Anne & Lynn Wheeler writes:

sum(IQ0, IQ1, ..., IQn)
sum(IQ0, IQ1, ..., IQn)/n
min(IQ0, IQ1, ..., IQn)
min(IQ0, IQ1, ..., IQn)/n

i.e. a comittee can be less effective than any single member of the
comittee alone ... with large comittees tending to zero effectiveness.

or taking a leaf out of management theory & span of control
(combinatorial interaction issue) ...

sum(IQ0, IQ1, ..., IQn)/n!

which also tends to zero for large n ...

it possibly is unrelated to the lowest IQ or even the inverse
... somebody once observed that effectiveness declined sharply as the
number of really bright (opinionated) people went up (something akin
to deadly-embrace scenario).

--
Anne & Lynn Wheeler   | lynn@garlic.com -  http://www.garlic.com/~lynn/

Why did OSI fail compared with TCP-IP?

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Why did OSI fail compared with TCP-IP?
Newsgroups: comp.arch,comp.protocols.iso,comp.protocols.tcp-ip,comp.lang.c++
Date: Sat, 04 May 2002 00:22:54 GMT

Andi Kleen writes:

TCP/IP does not really conform to the OSI 7 layer model, although
it is not too far off.

except for the whole IP (internet) layer ... which is non-existant in
the OSI model ... and would create a brand new 8th layer ... sort of
sitting somewhere between transport (layer 4) and network (layer 3).

the IP (internet) layer ... was the major thing in the 1/1/83 great
switch-over. the pre-IP, pre-1/1/83, NCP/IMP based infrastructure was
not too far off (from OSI model), being much more of a traditional,
homogeneous networking implementation. The 1/1/83 switch-over to an
internetworking implementation added something that was different with
the (IP) layer to inter-networking of networks (networks being a layer
3 concept, and inter-networking not having an OSI defintion).

The "inter-network" IP layer ... also including the concept of
(inter)-networking gateways. My claim previous claim was that the
internal corporate network effectively had gateway function support in
every node and was one of the reasons that the internal corporate
network was larger than the whole arpa/internet until sometime circa
1985. I claim that the introduction of the IP layer and
internetworking with gateways was one of the things that help promote
the take-off of the internet (in 1/1/83 switch-over) and for it to
surpose the internal network in size (there was a number of other
factors also).

misc. past 1/1/83 switch over postings
http://www.garlic.com/~lynn/2001c.html#4 what makes a cpu fast
http://www.garlic.com/~lynn/2001e.html#16 Pre ARPAnet email?
http://www.garlic.com/~lynn/2001l.html#35 Processor Modes
http://www.garlic.com/~lynn/2001m.html#48 Author seeks help - net in 1981
http://www.garlic.com/~lynn/2001m.html#54 Author seeks help - net in 1981
http://www.garlic.com/~lynn/2001n.html#6 Author seeks help - net in 1981
http://www.garlic.com/~lynn/2001n.html#87 A new forum is up! Q: what means nntp
http://www.garlic.com/~lynn/2002.html#32 Buffer overflow
http://www.garlic.com/~lynn/2002b.html#53 Computer Naming Conventions
http://www.garlic.com/~lynn/2002b.html#58 ibm vnet : Computer Naming Conventions

--
Anne & Lynn Wheeler   | lynn@garlic.com -  http://www.garlic.com/~lynn/

The need for Meaningful Error Messages :)

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: The need for Meaningful Error Messages :)
Newsgroups: alt.folklore.computers
Date: Sun, 05 May 2002 23:14:54 GMT

ab528@FreeNet.Carleton.CA (Heinz W. Wiggeshoff) writes:

And that's the REAL problem.  The quicker the computing world
rewinds to, say, 1975 fer instance, the better.  Give me a box
of punched cards, an 029 and I'll be back in my 'happy place'.
Oh - throw in a 2741 APL terminal too.

or at least APL typeball and a system that has APL 2741 support. I
don't have my 2741 anymore (almost 30 years gone ... although I only
got around to throwing out the table-top cover that fit on my 2741
just a couple years ago), but I do have one thing left which is a 2741
APL typeball. random apl refs:
http://www.garlic.com/~lynn/subtopic.html#hone

--
Anne & Lynn Wheeler   | lynn@garlic.com -  http://www.garlic.com/~lynn/

Security Issues of using Internet Banking

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Security Issues of using Internet Banking
Newsgroups: alt.computer.security,alt.folklore.computers
Date: Sun, 05 May 2002 23:21:58 GMT

somewhat related to internet banking security ... i updated my merged
security glossary with NSA's intrustion glossary. payment, financial,
x9f, and security glossaries at:
http://www.garlic.com/~lynn/index.html#glossary

the addition of this glossary along with some associated references
seems to beg for new taxonomy structure for "security software" (I
added some ... but not extensive).
http://www.garlic.com/~lynn/secure.htm

also somewhat related to secure financial transactions is the
non-repudiation thread on PKIX mailing list:
http://www.garlic.com/~lynn/aadsm11.htm#5 Meaning of Non-repudiation
http://www.garlic.com/~lynn/aadsm11.htm#6 Meaning of Non-repudiation
http://www.garlic.com/~lynn/aadsm11.htm#7 Meaning of Non-repudiation

--
Anne & Lynn Wheeler   | lynn@garlic.com -  http://www.garlic.com/~lynn/

Why is DSA so complicated?

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Why is DSA so complicated?
Newsgroups: sci.crypt
Date: Mon, 06 May 2002 02:53:05 GMT

ashwood@msn.com (Joseph Ashwood) writes:

Additionally since DSA can be performed using ECC (ECDSA), this gives
the possibility of some very small keys, making it ideally suited for
small devices.

I think that covers most of the reasons why DSA, for all it's seeming
complexity, is often preferrable to RSA-style signatures. Of course
I'm sure that given provocation I could come up with a sizable list
why RSA signatures are superior as well.

one of the issues with DSA (both fips186-1 & fips182-2) has been
secure random number generation. while ec/dsa looks good from the
standpoint of small devices ... chipcards, etc ... many of these
devices in the past have had inadequate random number capability.  In
the past I heard of some reference to test of various older model
chipcards ... involving power-on, generate random number, power-off
... repeated 64k times ... and there were repeats involving possibly
up to 30 percent of the generated numbers.

As a result, they have tended to use RSA keys ... where the keys have
been generated externally and injected into the device ... and then
used to sign messages ... which have had random numbers included in
the body of the message before passing to the device (NONCEs).

Given a "small device" with adequate random number capability ... then
it is possible to do both DSA (ECC or otherwise) key generation in the
device as well as DSA signatures (requiring random number generated
for each signature). Had an opportunity to have booth demonstrating
one such hardware token at cardtech/securetech two weeks ago in New
Orleans. misc refs to aads chip strawman:
http://www.garlic.com/~lynn/x959.html#aads

NIST FIPS page:
http://csrc.nist.gov/publications/fips/index.html

... with respect to fips186-2:

DRAFT
October 2001 -- A change notice for FIPS 186-2, Digital Signature
Standard (DSS) (.pdf file), has been made available that addresses key
sizes and random number generation. This change notice replaces the
item that was posted on August 3, 2001, Recommendations Regarding
Federal Information Processing Standard (FIPS) 186-2, Digital
Signature Standard (DSS). Comments and questions for this
recommendation are requested and may be addressed to FIPS186@nist.gov.

... from fips186-1:

 APPENDIX 3. RANDOM NUMBER GENERATION FOR THE DSA

Any implementation of the DSA requires the ability to generate random
or pseudorandom integers.  Such numbers are used to derive a user's
private key, x, and a user's per message secret number, k.  These
randomly or pseudorandomly generated integers are selected to be
between 0 and the 160-bit prime q (as specified in the standard).
They shall be generated by the techniques given in this appendix, or
using other FIPS approved security methods.

One FIPS approved pseudorandom integer generator is supplied in
Appendix C of ANSI X9.17, "Financial Institution Key Management
(Wholesale)."

--
Anne & Lynn Wheeler   | lynn@garlic.com -  http://www.garlic.com/~lynn/

"Soul of a New Machine" Computer?

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: "Soul of a New Machine" Computer?
Newsgroups: alt.folklore.computers
Date: Mon, 06 May 2002 12:41:58 GMT

Eric Smith <eric-no-spam-for-me@brouhaha.com> writes:

The first 88K processor, the 88100, was even externally a Harvard
architecture.  It used separate 88200 CMMU (Cache & MMU) chips on the
instruction bus and the data bus.  Normally the system bus side of the
CMMUs would be tied together, but there was no reason the hardware required
that.

IIRC, the AMD 29000 and 29005 used separate instruction and data busses
as well, but a single common address bus.

i don't know about 88k, but engineers that had worked on 801 designs
were later involved in working on both amd 29k and hp's risc. I don't
remember the exact timing ... but it may have been after Fort Knox was
killed that you started seeing 801 engineers at other companies. misc.
fort knox references:
http://www.garlic.com/~lynn/99.html#136a checks (was S/390 on PowerPC?)
http://www.garlic.com/~lynn/2000d.html#60 "all-out" vs less aggressive designs (was: Re: 36 to 32 bit transition)
http://www.garlic.com/~lynn/2001f.html#43 Golden Era of Compilers
http://www.garlic.com/~lynn/2001h.html#69 Very CISC Instuctions (Was: why the machine word size ...)

with regard to:
http://www.garlic.com/~lynn/2002g.html#18 Black magic in POWER5

there is folklore that the individual worked last two weeks up until
final hour on blue iliad (which never got much past early sample
chips; romp was 16bit chip, blue iliad was first 32bit, big & hot; &
RIOS was 32bit):
http://www.garlic.com/~lynn/98.html#25 Merced & compilers (was Re: Effect of speed ... )
http://www.garlic.com/~lynn/99.html#66 System/1 ?
http://www.garlic.com/~lynn/2000.html#16 Computer of the century
http://www.garlic.com/~lynn/2000d.html#60 "all-out" vs less aggressive designs (was: Re: 36 to 32 bit transition)

random romp postings:
http://www.garlic.com/~lynn/98.html#26 Merced & compilers (was Re: Effect of speed ... )
http://www.garlic.com/~lynn/98.html#27 Merced & compilers (was Re: Effect of speed ... )
http://www.garlic.com/~lynn/99.html#64 Old naked woman ASCII art
http://www.garlic.com/~lynn/99.html#129 High Performance PowerPC
http://www.garlic.com/~lynn/2000.html#49 IBM RT PC (was Re: What does AT stand for ?)
http://www.garlic.com/~lynn/2000.html#59 Multithreading underlies new development paradigm
http://www.garlic.com/~lynn/2000b.html#54 Multics dual-page-size scheme
http://www.garlic.com/~lynn/2000c.html#4 TF-1
http://www.garlic.com/~lynn/2001c.html#84 database (or b-tree) page sizes
http://www.garlic.com/~lynn/2001d.html#12 database (or b-tree) page sizes
http://www.garlic.com/~lynn/2001f.html#22 Early AIX including AIX/370
http://www.garlic.com/~lynn/2001f.html#43 Golden Era of Compilers
http://www.garlic.com/~lynn/2002c.html#40 using >=4GB of memory on a 32-bit processor
http://www.garlic.com/~lynn/2002g.html#17 Black magic in POWER5

--
Anne & Lynn Wheeler   | lynn@garlic.com -  http://www.garlic.com/~lynn/

Why did OSI fail compared with TCP-IP?

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Why did OSI fail compared with TCP-IP?
Newsgroups: comp.arch,comp.protocols.iso,comp.protocols.tcp-ip
Date: Mon, 06 May 2002 13:10:11 GMT

"A. G. McDowell" writes:

To be somewhat more constructive, I think there are so many ways to get
nowhere that the more useful question is "How did DARPA do so well?".
Many of those scarred by OSI were also in touch with TCP/IP and did
learn from it. Here's to "Rough Consensus and Running Code"!

Internop '88 had a large number of vendors that were also showing OSI
related products. Interop '88 was possibly the first time that a large
number of different boxes were connected to multiple LANs ... the
floor NET was four parallel LANs (remember OSI believed that LANs were
invalid and shouldn't be allowed to exist). Sunday the floor nets were
crashing welling into monday AM ... before the problem was diagnosed
... which resulted in a new (IETF) standard specification.

the concept of interneting & gateways (again something not provided
for by OSI and was ruled invalid/violations) had come about 1/1/83 and
in part allowed the internet to exceed the size of the internal
corporate network by sometime '85 (pre 1/1/83, non-internetworking &
non-gateway was much more straight OSI).

NSF had let RFP for NSFNET1 backbone ... however it came about at time
that the internet was growing, interneting/gateways was prooving valid
and there was huge excessive dark fiber capacity sitting around. The
result was that the amount of commercial resources dumped into the
backbone was far in excess of the funding provided by the NSF NSFNET
(& NSFNET2) backbone RFP (direct & indirect commercial backing was far
in excess of federal funding; there has been side threads in the past
about the acceptable use policies not allowing commercial use of the
NSFNET backbone because it was total federal backing). A conjecture
was that the dark fiber owners were attempting to promote both actual
use and the academic development of applications oriented towards
large bandwidth utilization (planting seeds for utilization demand of
the dark fiber bandwidth). One could say that they succeeded ... most
notably with the advent of Mosaic.

random nsfnet & interop refs:
http://www.garlic.com/~lynn/internet.htm
http://www.garlic.com/~lynn/94.html#34 Failover and MAC addresses (was: Re: Dual-p
http://www.garlic.com/~lynn/94.html#36 Failover and MAC addresses (was: Re: Dual-p
http://www.garlic.com/~lynn/98.html#49 Edsger Dijkstra: the blackest week of his professional life
http://www.garlic.com/~lynn/98.html#59 Ok Computer
http://www.garlic.com/~lynn/99.html#33 why is there an "@" key?
http://www.garlic.com/~lynn/99.html#37a Internet and/or ARPANET?
http://www.garlic.com/~lynn/99.html#37b Internet and/or ARPANET?
http://www.garlic.com/~lynn/99.html#38c Internet and/or ARPANET?
http://www.garlic.com/~lynn/99.html#40 [netz] History and vision for the future of Internet - Public Question
http://www.garlic.com/~lynn/99.html#138 Dispute about Internet's origins
http://www.garlic.com/~lynn/99.html#146 Dispute about Internet's origins
http://www.garlic.com/~lynn/subtopic.html#subject Postings by various subjects
http://www.garlic.com/~lynn/2000.html#49 IBM RT PC (was Re: What does AT stand for ?)
http://www.garlic.com/~lynn/2000c.html#26 The first "internet" companies?
http://www.garlic.com/~lynn/2000c.html#59 Does the word "mainframe" still have a meaning?
http://www.garlic.com/~lynn/2000c.html#78 Free RT monitors/keyboards
http://www.garlic.com/~lynn/2000d.html#16 The author Ronda Hauben fights for our freedom.
http://www.garlic.com/~lynn/2000d.html#19 Comrade Ronda vs. the Capitalist Netmongers
http://www.garlic.com/~lynn/2000d.html#43 Al Gore: Inventing the Internet...
http://www.garlic.com/~lynn/2000d.html#56 Is Al Gore The Father of the Internet?
http://www.garlic.com/~lynn/2000d.html#58 Is Al Gore The Father of the Internet?
http://www.garlic.com/~lynn/2000d.html#59 Is Al Gore The Father of the Internet?
http://www.garlic.com/~lynn/2000d.html#63 Is Al Gore The Father of the Internet?
http://www.garlic.com/~lynn/2000d.html#70 When the Internet went private
http://www.garlic.com/~lynn/2000d.html#71 When the Internet went private
http://www.garlic.com/~lynn/2000d.html#72 When the Internet went private
http://www.garlic.com/~lynn/2000d.html#73 When the Internet went private
http://www.garlic.com/~lynn/2000d.html#74 When the Internet went private
http://www.garlic.com/~lynn/2000d.html#77 Is Al Gore The Father of the Internet?^
http://www.garlic.com/~lynn/2000e.html#5 Is Al Gore The Father of the Internet?^
http://www.garlic.com/~lynn/2000e.html#10 Is Al Gore The Father of the Internet?^
http://www.garlic.com/~lynn/2000e.html#11 Is Al Gore The Father of the Internet?^
http://www.garlic.com/~lynn/2000e.html#19 Is Al Gore The Father of the Internet?^
http://www.garlic.com/~lynn/2000e.html#28 Is Al Gore The Father of the Internet?^
http://www.garlic.com/~lynn/2000e.html#29 Vint Cerf and Robert Kahn and their political opinions
http://www.garlic.com/~lynn/2000e.html#31 Cerf et.al. didn't agree with Gore's claim of initiative.
http://www.garlic.com/~lynn/2000f.html#44 Al Gore and the Internet (Part 2 of 2)
http://www.garlic.com/~lynn/2000f.html#47 Al Gore and the Internet (Part 2 of 2)
http://www.garlic.com/~lynn/2000f.html#50 Al Gore and the Internet (Part 2 of 2)
http://www.garlic.com/~lynn/2000f.html#51 Al Gore and the Internet (Part 2 of 2)
http://www.garlic.com/~lynn/2001.html#4 Sv: First video terminal?
http://www.garlic.com/~lynn/2001d.html#42 IBM was/is: Imitation...
http://www.garlic.com/~lynn/2001e.html#76 Stoopidest Hardware Repair Call?
http://www.garlic.com/~lynn/2001h.html#44 Wired News :The Grid: The Next-Gen Internet?
http://www.garlic.com/~lynn/2001h.html#74 YKYGOW...
http://www.garlic.com/~lynn/2001i.html#5 YKYGOW...
http://www.garlic.com/~lynn/2001i.html#6 YKYGOW...
http://www.garlic.com/~lynn/2002.html#33 Buffer overflow

--
Anne & Lynn Wheeler   | lynn@garlic.com -  http://www.garlic.com/~lynn/

Why is DSA so complicated?

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Why is DSA so complicated?
Newsgroups: sci.crypt
Date: Mon, 06 May 2002 18:21:40 GMT

"Roger Schlafly" writes:

There is not much difference regarding random numbers. The consensus
now is that you also need random numbers to do RSA securely. Either
way, a pseudorandom number generator could be injected into the device.

RSA is not really any simpler, either. Just look at all the papers on
the security of various padding schemes and the other RSA technicalities.
The advantage of RSA is that it has faster signature verifications, and
that's about all.

there is appearnce of difference of regarding requiring random
numbers.  It is obvious that a hardware token needs a secure random
number for generating the actual signatures.

there is a lot of stuff about padding and including random numbers in
the body of messages that are RSA signed ... however since it isn't
actually part of the specification ... they allow external forces to
supply all of that externally before sending the message to the token
for signing (aka random numbers all being done externally to the
token). that leaves the token dependent on external forces for the
random numbers & padding (possible choice when tokens were chosen w/o
real random number capability, keys generated externally, and just
SHA-1 sent to token for RSA signing).

This possibly exposes a token to risks of signing mal-formed
messages/SHA-1? ... which is mitigated when token accepts full
responsibility for fips186 operation (security issues of signing
dependencies are all encapsulated explicitly in the standard).

There is possibility that any belief that RSA was much simpler and
didn't have any of these other security issues allowed for bifurcated
implementations ... with tokens just doing simple RSA encrypting of
sha-1.

--
Anne & Lynn Wheeler   | lynn@garlic.com -  http://www.garlic.com/~lynn/

Why is DSA so complicated?

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Why is DSA so complicated?
Newsgroups: sci.crypt
Date: Mon, 06 May 2002 18:21:40 GMT

"Roger Schlafly" writes:

There is not much difference regarding random numbers. The consensus
now is that you also need random numbers to do RSA securely. Either
way, a pseudorandom number generator could be injected into the device.

RSA is not really any simpler, either. Just look at all the papers on
the security of various padding schemes and the other RSA technicalities.
The advantage of RSA is that it has faster signature verifications, and
that's about all.

there is appearnce of difference of regarding requiring random
numbers.  It is obvious that a hardware token needs a secure random
number for generating the actual signatures.

there is a lot of stuff about padding and including random numbers in
the body of messages that are RSA signed ... however since it isn't
actually part of the specification ... they allow external forces to
supply all of that externally before sending the message to the token
for signing (aka random numbers all being done externally to the
token). that leaves the token dependent on external forces for the
random numbers & padding (possible choice when tokens were chosen w/o
real random number capability, keys generated externally, and just
SHA-1 sent to token for RSA signing).

This possibly exposes a token to risks of signing mal-formed
messages/SHA-1? ... which is mitigated when token accepts full
responsibility for fips186 operation (security issues of signing
dependencies are all encapsulated explicitly in the standard).

There is possibility that any belief that RSA was much simpler and
didn't have any of these other security issues allowed for bifurcated
implementations ... with tokens just doing simple RSA encrypting of
sha-1.

--
Anne & Lynn Wheeler   | lynn@garlic.com -  http://www.garlic.com/~lynn/

Why is DSA so complicated?

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Why is DSA so complicated?
Newsgroups: sci.crypt
Date: Mon, 06 May 2002 19:36:42 GMT

"Anton Stiglic" writes:

What do you mean it's not part of the specification?

of what gets implemented in a hardware token that is passed a sha-1 to
encrypted with RSA private key ... there is nothing about the hardware
token being able to verify that it isn't dealing with a mal-formed
message and refusing to sign it.

lets say where is the trust/security boundary ... is the
trust/security boundary only around the physical perimeter of the chip
inside the hardware token ... or does the trust/security boundary
extend to all the components that the chip might be dependent on for
correct operation.

--
Anne & Lynn Wheeler   | lynn@garlic.com -  http://www.garlic.com/~lynn/

ibm icecube -- return of watercooling?

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: ibm icecube -- return of watercooling?
Newsgroups: comp.arch,comp.arch.storage
Date: Tue, 07 May 2002 04:14:24 GMT

"del cecchi" writes:

The IBM 4381 was impingement cooled.  :-)

it was more like an upright freezer (aka rack) than the precursor 4341
which was more like a chest freezer. there all the jokes about the
4381 having this little "cooling towers" inside.

quick check with alta-vista just turned up these pictures:
http://www.telnet.hu/hamster/oldiron/e_ibms.html

didn't show insides of the 4381 ... but did show inside 3090

--
Anne & Lynn Wheeler   | lynn@garlic.com -  http://www.garlic.com/~lynn/

Why did OSI fail compared with TCP-IP?

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Why did OSI fail compared with TCP-IP?
Newsgroups: comp.arch,comp.protocols.iso,comp.protocols.tcp-ip
Date: Tue, 07 May 2002 15:42:16 GMT

craigp@world.std.com (Craig Partridge) writes:

There was not a huge excess of capacity for NSFNET Phase I,
indeed, it was a struggle to get T1 lines for some locations.  Nor was
there much corporate contribution.  NSFNET Phase I was run on a shoestring.

I believe we have agreed offline that I was referring to NSFNET1
(phase I) & NSFNET2 (phase 2) backbone RFP for T1 & T3 backbones.

What I was calling NSFNET1 backbone ... was the RFP for "T1" backbone
which was "won" by Merit, IBM, & MCI (for $11.2m, if I remember
right), and NSFNET2 backbone was for RFP for "T3" backbone. Both the
NSFNET1 backbone and NSFNET2 backbone were heavily subsidized by
commercial sources far in excess of the RFP bid awards/funding by the
federal gov:
http://www.garlic.com/~lynn/2000e.html#10

Earlier NSFNET/CSNET efforts funded by NSFNET were shoestring (aka not
the heavy commerical subsidy of later activities) ... somewhat
related:
http://www.garlic.com/~lynn/99.html#37a

only slightly related:
http://www.garlic.com/~lynn/2001h.html#65

--
Anne & Lynn Wheeler   | lynn@garlic.com -  http://www.garlic.com/~lynn/

Why did OSI fail compared with TCP-IP?

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Why did OSI fail compared with TCP-IP?
Newsgroups: comp.arch,comp.protocols.iso,comp.protocols.tcp-ip,alt.folklore.computers
Date: Tue, 07 May 2002 16:02:21 GMT

hfbonney writes:

Well, MAP and TOP were ISO stacks over 802.4 token bus and over ethernet,
respectively. MAP (Manufacturing Automation Protocol) added things
like mini-MAP where devices were connected at Layer 2 within a
manufacturing "cell". There must have been more implementation work
done on this standard for quite a while. I think Boeing went for it,
for one, though it was GM that started it (?). Does it still exist at
all?

but was it an ISO OSI stack?

ANSI X3S3.3 had standards responsibility for approx. network/transport
level responsibility and fed into the corresponding standards body at
the ISO level. the message was that x3s3.3 could work on HSP (high
speed protocol) but that the ISO group wouldn't pass it as a standard
... because the charter for that standards group was that standards
had to conform to the OSI model (and the ISO group responsible for
lower level standards activity also had charter that stated they could
only pass standards that conformed to the OSI model).

That doesn't say that there weren't other ANSI and ISO groups that
could pass standards ... which didn't require conforming to OSI model
... especially if they were in specialized industries.

--
Anne & Lynn Wheeler   | lynn@garlic.com -  http://www.garlic.com/~lynn/

Why are Mainframe Computers really still in use at all?

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Why are Mainframe Computers really still in use at all?
Newsgroups: alt.folklore.computers
Date: Tue, 07 May 2002 16:13:57 GMT

"Russell P. Holsclaw" writes:

So, to run PCP on a model 65 was absurd in the extreme.  I'm pretty sure the
IRS box had at least 512K (that seemed like a lot in those days :-) ),
although I believe it was possible to buy a mod65 with only 256K, but no
less. The core arrays came in 256K increments, IIRC. In fact, much of the
speed advantage of the 65 came from interleaved core cycles, which probably
required more than the 256K configuration.

unless they might have had a dedicated application that took all of
the machine all of the time ... I know one customer with such an
application ... they were eventually resorting to tweaking DOS so that
it would run on 370/165 (pre-relocate).

360/50 had 2mic memory ... 360/60 & 360/70 were to have 1mic memory
... before first customer ship, memory technology was upgraded to
750ns memory and the model numbers changed to 360/65 & 360/75. There
were some 360/65s early-on running with 256k ... but operating system
requirements eventually forced customers to larger memory
configurations.


note/update:

I remember reading an early document about 360/6x machine with virtual
memory having one, two, and four processors. I sort of had vaque
recollection that it was model number other than 360/67.

however, i've subsequently been told that 360/60 was with 2mic memory
and 360/62 was with 1mic memory. both models never shipped, and were
replaced with 360/65 with 750ns memory. the 360/67 then shipped as
360/65 with virtual memory ... only available in one (uniprocessor)
and two processor (multiprocessor) configurations


--
Anne & Lynn Wheeler   | lynn@garlic.com -  http://www.garlic.com/~lynn/

Why did OSI fail compared with TCP-IP?

From: Anne & Lynn Wheeler <lynn@garlic.com>
Subject: Re: Why did OSI fail compared with TCP-IP?
Newsgroups: comp.arch,comp.protocols.iso,comp.protocols.tcp-ip,comp.lang.c++
Date: Tue, 07 May 2002 17:58:18 GMT

"Rudvar Alswill" writes:

The layered approach, really nothing to shout about, was not a OSI
achievement. They borrowed this from IBMs SNA and Honeywells network.
OSI contribution seems almost negative, giving people th