Latest IBM Reference Card?
Lisa or Jeff
At work, I use an old IBM S/370 reference card, which is adequate for my
purposes. Mine is a yellow card, which replaced a green S/360 card.
I understand even in S/370 days, they added so much stuff they replaced
the card with a booklet.
I was wondering if there is still a "quick reference" card (or booklet),
and if so, what's in today's references.
And, if it sill carrieds punch card codings?
Bob Shair
hanc...@bbs.cpcn.com (Lisa or Jeff) writes:
>At work, I use an old IBM S/370 reference card, which is adequate for my
>purposes. Mine is a yellow card, which replaced a green S/360 card.
>
>I understand even in S/370 days, they added so much stuff they replaced
>the card with a booklet.
>
>I was wondering if there is still a "quick reference" card (or booklet),
ORDERNO GX20-1850 SUFFIX: 07
TITLE S/370 REFERENCE SUMMARY
DESCRIBE PAPER, 8.500" X 3.750", 44 PAGES
UNIT/MEASURE 001
DISTRIB BILLABLE SLSS AVAILABLE
PRICE SELL PRICE: $1.95
DATE AVAILABLE DATE: 04/72 LATEST DASH LEVEL 03/89
>and if so, what's in today's references.
Mine is suffix -6, and contains machine opcodes, instruction formats,
condition codes, assemmbler instructions, control register, fixed storage,
and PSW layouts, interruption codes, DAT page and segment table layouts,
CCW, CSW, and CAW layouts, plus list of channel commands for card readers,
printers, tapes and disk drives. BSC control characters, and (of course)
code point assignments for BCD, EBCDIC, ASCII and
(of course)
punch cards.
>
>And, if it sill carrieds punch card codings?
No paper tape, though.
Those of us with newer systems should probably use, instead,
ORDERNO SA22-7209 SUFFIX: 01
TITLE IBM Enterprise Systems Architecture/390 Reference Summary
ABSTRACT This publication is intended primarily for use by Enterprise
Systems Architecture/390 assembler language application
programmers. It contains basic machine information
summarized from the IBM Enterprise Systems Architecture/390
Principles of Operation, SA22-7201.
DESCRIBE PAPER, 8.500" X 3.630", 64 PAGES
UNIT/MEASURE 001
DISTRIB BILLABLE SLSS AVAILABLE
PRICE SELL PRICE: $15.25
DATE AVAILABLE DATE: 07/94
but who could afford it?
--
Bob Shair Open Systems Consultant
1018 W. Springfield Avenue rms...@uiuc.edu
Champaign, IL 61821 217/356-2684
< Not employed by or representing the University of Illinois >
gl...@glass.cv.lexington.ibm.com
Actually, the sequence went from the green card (S/360, Order number
GX20-1703-9), to the yellow card (S/370, GX20-1850), to the yellow
booklet (S/370, GX20-1850-7), to the pink booklet (Ok, 'Salmon coloured')
(S/370 Extended Architecture, GX20-0157-2), to the blue booklet (ESA/370,
GX20-0406-0), to the white booklet (ESA/390, SA22-7209-01).
I suspect I've left out a few (Wasn't there a white card in the S/360 era
for a specific machine?). And, for the collectors out there, I'm told
that GX20-0157-0, the first S/370 Extended Architecture Reference Summary
is a bit of a collector's item (Due to the way the name was printed on the
cover, it was ocassionally referred to as the 'SEARS' card). :*)
Dave
P.S. Standard Disclaimer: I work for them, but I don't speak for them.
Joel C. Ewing
Lisa or Jeff wrote:
>
> At work, I use an old IBM S/370 reference card, which is adequate for my
> purposes. Mine is a yellow card, which replaced a green S/360 card.
>
> I understand even in S/370 days, they added so much stuff they replaced
> the card with a booklet.
>
> I was wondering if there is still a "quick reference" card (or booklet),
>
I believe the transition went to a yellow booklet, next to the pink
booklet (GX20-0157 for XA architecture, 45 pgs), and then to a pastel
blue booklet for ESA architecture (I have one at work but can't remember
the number). They are still called reference summaries and they still
contain a character code table (EBCDIC and ASCII) with punch card
representations, all the machine instructions and assembler codes, plus
other assorted info. The greatest changes have mostly occurred in areas
which are of marginal interest to application programmers, so some of
the older reference summaries are still quite useful.
--
Joel C. Ewing, Fort Smith, AR jce...@acm.org
Phil Smith III
ObAnecdote: a friend at a large NY brokerage house was working a problem with
one of the junior people, a non-native of the US. At some point he wanted to
check something about an instruction, and asked for her green card. With a
*very* odd look, she reached into her wallet and handed him her INS card...
sidney orr
The moral or theme of this exchange might be that there was no doubt that
IBM was for many years and still is perhaps the premier _technology_
company in many ways; but, when your management gets up to nearly two-dozen
layers, none of whom could even spell i n t e r n e t until 1994 or
so (despite having _built_ much of it) the enterprise is in trouble, no
matter how good its technology is... appropros big blue's technology
I recommend <Computer Wars: The Fall of IBM And The Rise Of Global
Technology> - still in print...
--
John Savard
gl...@glass.cv.lexington.ibm.com wrote:
>IBM Enterprise Systems Architecture/390 Principles of Operation, SA22-7201.
Ah! The local University bookstore doesn't carry these anymore...but
now I know the number, and can try ordering it from IBM (sheesh...if
the reference card is $15.25, this, of course, will be more
expensive...) to keep up-to-date.
John Savard
Bill Donald
I purchased the latest version the POO book just before Christmas,
as it's ancient predecessor had literally fallen apart. ISTR it
costing around 12 UK pounds, so I'd expect the native price to
be around $18. Not bad value for a book the size of your average
Yellow Pages.
IBM reference cards are ALWAYS expensive, relative to their scope
and it's often more cost-effective to buy the full size book.
Another example that springs to mind is the Bookmaster Quick
Reference booklet and the User Guide - one is the size of a
birthday card, the other is a BT phone book.
--------------------------------------------------------------------
Bill Donald G7ENQ
Euro Interconnect Signaling Systems (44) 01628 431063
Nortel Ltd. Maidenhead ENGLAND. home: wdo...@ibm.net
"What I'm out for's a good time. All the rest is propaganda"
Alan Sillitoe 'Saturday Night and Sunday Morning' 1960.
--------------------------------------------------------------------
Andy Styles
Bill Donald wrote:
> I purchased the latest version the POO book just before Christmas,
> as it's ancient predecessor had literally fallen apart. ISTR it
> costing around 12 UK pounds, so I'd expect the native price to
> be around $18. Not bad value for a book the size of your average
> Yellow Pages.
I've just bought the ESA 390 POp - it cost me just under 12 GBPs I
think. I was stunned, I seem to remember the POp costing stacks a few
years ago!
I must get the HLASM Reference and the MVS ESA Services Reference now!
--
Andy Styles.
My opinions are exclusively my own. No-one else wants them. (<sniff>)
John Wilson
In article <332738...@ibmmail.com>,
Andy Styles <usfm...@ibmmail.com> wrote:
>I've just bought the ESA 390 POp - it cost me just under 12 GBPs I
>think. I was stunned, I seem to remember the POp costing stacks a few
>years ago!
Well at least for the last decade or so they've been way cheap, if I remember
right I paid $2.99 for the 360 POOP in 1984 (but it's a lot thinner than the
370 POOP), and got a couple of versions of the 370 POOP for around $15 after
that. IIRC the 360 POOP has a bunch of charts for various codes in the back,
BCDIC, EBCDIC, 029 etc. and then they list ASCII "for completeness", with a
footnote saying that no IBM product uses ASCII (funny I thought the 360 still
had an ASCII mode) but if one ever does, they will replace the "^" character
with a funny weather symbol, or some such bullshit. I love it! Leave no
standard unviolated... AFAIK it was a lie, the 3101 and 5150 use perfectly
normal ASCII for printing characters anyway.
John Wilson
D Bit
Chuck Greenwald
In article <5ftjhk$s...@tor-nn1-hb0.netcom.ca>, sew...@netcom.ca (John Savard) writes:
|> gl...@glass.cv.lexington.ibm.com wrote:
|>
|> >IBM Enterprise Systems Architecture/390 Principles of Operation, SA22-7201.
|>
|> Ah! The local University bookstore doesn't carry these anymore...but
|> now I know the number, and can try ordering it from IBM (sheesh...if
|> the reference card is $15.25, this, of course, will be more
|> expensive...) to keep up-to-date.
The P. of O. is close to a thousand pages of fine print.
The reference card is:
Enterprise Systems Architecture/390 Reference Summary, SA22-7209.
(Up to 55 pages now -- I can't find my green card, but the yellow one
from '76 was 8 pannels.)
------------------------------------------------------------------------
Chuck Greenwald | All opinions are | chu...@meaddata.com
Lexis/Nexis | my own. | (513) 865-1020
Eric Jackson
John Wilson wrote:
>
> In article <332738...@ibmmail.com>,
> Andy Styles <usfm...@ibmmail.com> wrote:
> >I've just bought the ESA 390 POp - it cost me just under 12 GBPs I
> BCDIC, EBCDIC, 029 etc. and then they list ASCII "for completeness", with a
> footnote saying that no IBM product uses ASCII (funny I thought the 360 still
> had an ASCII mode) but if one ever does, they will replace the "^" character
> with a funny weather symbol, or some such bullshit. I love it! Leave no
> standard unviolated...
The 360 had an ASCII mode, but the only effect was to change the sign
generated by decimal operations. No IBM operating system ran with that
bit turned on, and the PSW bit was reused in the 370 for virtual mode.
> AFAIK it was a lie, the 3101 and 5150 use perfectly
> normal ASCII for printing characters anyway.
>
> John Wilson
> D Bit
When the 360 POO was written, the 3101 and 5150 were not products. When
they wrote it, that's what they intended to happen if the government
forced ASCII on them. Instead it was market forces that pushed them
into ASCII. There's a very interesting book in the IBM DP series about
the history of character codes called _Character Coded Sets_ or
something like that.
John Starling
Chuck Greenwald (chu...@lexis-nexis.com) wrote:
: The P. of O. is close to a thousand pages of fine print.
: The reference card is:
: Enterprise Systems Architecture/390 Reference Summary, SA22-7209.
: (Up to 55 pages now -- I can't find my green card, but the yellow one
: from '76 was 8 pannels.)
sometimes, you will simply avoid using the appropriate instruction.
If you work where they hold IBM licences, you should have access
to the CD-ROM image for the doc under book manager. Free? Who has
checked into this? Book manager on net is not too good, but
I tried book manager on TSO/ISPF and that was very bad.
B.T.W. The oldest greenie I have is GX20-1703-7, so there would be
at least 6 versions before that. This one is 6 "panels", 12 pages.
No copyright date is shown. Is it c. 1968?
-- __ __
Thanx. --John W. Starling x-4733 /_/\/_/\ star...@lexis-nexis.com
:-) Search Service Development \_\/_/\/ Office: B5/F4/MA/54S05
The opinions I assert are opinion. \_\_\/\ (513) 865-6800 x-4733
Home: 682 Deerfield Dr. /_/_/\/ Lexis-Nexis, POB 933
Harrison, OH 45030 (513) 367-5653 \_\_\/ Dayton, OH 45401
bbrey...@aol.com
In article <5g7163$k...@mailgate.lexis-nexis.com>, chu...@lexis-nexis.com (Chuck Greenwald) writes:
>Subject: Re: Latest IBM Reference Card?
>From: chu...@lexis-nexis.com (Chuck Greenwald)
>Date: 12 Mar 1997 19:46:11 GMT
>
>In article <5ftjhk$s...@tor-nn1-hb0.netcom.ca>, sew...@netcom.ca (John Savard)
>writes:
>|> gl...@glass.cv.lexington.ibm.com wrote:
>|>
>|> >IBM Enterprise Systems Architecture/390 Principles of Operation,
>SA22-7201.
>|>
>|> Ah! The local University bookstore doesn't carry these anymore...but
>|> now I know the number, and can try ordering it from IBM (sheesh...if
>|> the reference card is $15.25, this, of course, will be more
>|> expensive...) to keep up-to-date.
>
>The P. of O. is close to a thousand pages of fine print.
>
>The reference card is:
>
>Enterprise Systems Architecture/390 Reference Summary, SA22-7209.
>
>(Up to 55 pages now -- I can't find my green card, but the yellow one
>from '76 was 8 pannels.)
>
>------------------------------------------------------------------------
>Chuck Greenwald | All opinions are | chu...@meaddata.com
>Lexis/Nexis | my own. | (513) 865-1020
>
>
>
C. 1968, the green card was sold at the Rutgers University Bookstore for $.18.
Bruce B. Reynolds, Systems Consultant: Founder of Trailing Edge Technologies---It May Not Be Broken Now, But How About the Year 2000??
gl...@glass.cv.lexington.ibm.com
In <5g90jh$6...@mailgate.lexis-nexis.com>, star...@meaddata.com (John Starling) writes:
>Chuck Greenwald (chu...@lexis-nexis.com) wrote:
>: The P. of O. is close to a thousand pages of fine print.
>: The reference card is:
>: Enterprise Systems Architecture/390 Reference Summary, SA22-7209.
>: (Up to 55 pages now -- I can't find my green card, but the yellow one
>: from '76 was 8 pannels.)
>sometimes, you will simply avoid using the appropriate instruction.
>
>If you work where they hold IBM licences, you should have access
>to the CD-ROM image for the doc under book manager. Free? Who has
>checked into this? Book manager on net is not too good, but
>I tried book manager on TSO/ISPF and that was very bad.
>
>B.T.W. The oldest greenie I have is GX20-1703-7, so there would be
>at least 6 versions before that. This one is 6 "panels", 12 pages.
>No copyright date is shown. Is it c. 1968?
>
>-- __ __
>Thanx. --John W. Starling x-4733 /_/\/_/\ star...@lexis-nexis.com
>:-) Search Service Development \_\/_/\/ Office: B5/F4/MA/54S05
>The opinions I assert are opinion. \_\_\/\ (513) 865-6800 x-4733
>Home: 682 Deerfield Dr. /_/_/\/ Lexis-Nexis, POB 933
>Harrison, OH 45030 (513) 367-5653 \_\_\/ Dayton, OH 45401
Mine's a -9 version (GX20-1703-9), and it was up to 14 pages
(7 panels) by then. Again, no copyright date, so I can't say
when it first appeared.
Dave
P.S. Oh, there's also an OS/2 version and a DOS version of
BookManager Read. I find that I like the DOS version better
than the TSO or CMS versions. And, the P-of-O is available on
some of the CDs, such as SK2T-2067-11, the VM Collection CD set.
P.P.S. Standard Disclaimer: I work for them, but I don't
speak for them.
Lisa or Jeff
> If you work where they hold IBM licences, you should have access
> to the CD-ROM image for the doc under book manager. Free? Who has
> checked into this? Book manager on net is not too good, but
> I tried book manager on TSO/ISPF and that was very bad.
Is the latest reference card for S/390 available on Book Manager? We have
the full Principles of Operation, but I was hoping to print down some
quick reference of the current instruction set. The Pr of Op is huge.
Edward Rice
In article <5g90jh$6...@mailgate.lexis-nexis.com>,
star...@meaddata.com (John Starling) wrote:
>B.T.W. The oldest greenie I have is GX20-1703-7, so there would be
>at least 6 versions before that. This one is 6 "panels", 12 pages.
>No copyright date is shown. Is it c. 1968?
I have an X20-1703-5 (no leading "G" designator), 12 panels. No date on
it, but because I can reckon by PPOE when I was likely to have been issued
it, I'd say it was definitely not 1967 or earlier; would very likely have
been 1969; could have been 1970.
Wayne L. Beavers
gl...@glass.cv.lexington.ibm.com wrote:
>
> Mine's a -9 version (GX20-1703-9), and it was up to 14 pages
> (7 panels) by then. Again, no copyright date, so I can't say
> when it first appeared.
>
> Dave
Mine's a - 9 also. I bought it at the Cal Poly (San Luis Obispo) book
store for 15 cents in 1972. We were running OS/PCP 19.6 at the time.
--
---
Wayne L. Beavers mailto:way...@beyond-software.com
Beyond Software, Inc. http://www.beyond-software.com
"The Mainframe/Internet Company"
James R. Harper
In article <5g932g$c...@ausnews.austin.ibm.com>
> P.S. Oh, there's also an OS/2 version and a DOS version of
> BookManager Read. I find that I like the DOS version better
> than the TSO or CMS versions. And, the P-of-O is available on
> some of the CDs, such as SK2T-2067-11, the VM Collection CD set.
>
you're using a SVGA monitor!! And if you've got a 3270 emulation software
on your PC you can download POPs, and most other IBM published manuals,
from the mainframe to your PC using INDFILE - but make sure you do it in
binary mode NOT text mode.
James
>
James Harper Upton-by-Chester, Chester, CH2 1EB, UK
ja...@harpers.demon.co.uk
http://www.harpers.demon.co.uk
Verne Arase
In article <332751...@candle.com>,
Eric Jackson <Eric_J...@candle.com> wrote:
>The 360 had an ASCII mode, but the only effect was to change the sign
>generated by decimal operations. No IBM operating system ran with that
>bit turned on, and the PSW bit was reused in the 370 for virtual mode.
I think it also affected UNPK, ED, and EDMK.
Verne Arase
>If you work where they hold IBM licences, you should have access
>to the CD-ROM image for the doc under book manager. Free? Who has
>checked into this? Book manager on net is not too good, but
>I tried book manager on TSO/ISPF and that was very bad.
The latest OS/2 BookManager is a schizophrenic descent into "the internet
is all" mentality. It actually translates BookManager format (which wasn't
that bad really) into HMTL, and forces you to use IBM's Web Explorer to
read it.
Even though connected to the internet by token ring, I still experience
freezes when it's up.
John Savard
Eric Jackson <Eric_J...@candle.com> wrote:
>John Wilson wrote:
>>
>> In article <332738...@ibmmail.com>,
>> Andy Styles <usfm...@ibmmail.com> wrote:
>> >I've just bought the ESA 390 POp - it cost me just under 12 GBPs I
>> BCDIC, EBCDIC, 029 etc. and then they list ASCII "for completeness", with a
>> footnote saying that no IBM product uses ASCII (funny I thought the 360 still
>> had an ASCII mode) but if one ever does, they will replace the "^" character
>> with a funny weather symbol, or some such bullshit. I love it! Leave no
>> standard unviolated...
At the time, none did.
Today, of course, the IBM PS/2 and the IBM PC, XT, and AT all use(d)
ASCII.
But the 'official' standard for ASCII to EBCDIC translation translates
ASCII ! instead of ASCII | to the EBCDIC vertical bar, and ASCII ^
instead of ASCII ~ to the logical NOT symbol. Apparently, getting PL/I
on ASR 33s was more important than the English language.
>The 360 had an ASCII mode, but the only effect was to change the sign
>generated by decimal operations. No IBM operating system ran with that
>bit turned on, and the PSW bit was reused in the 370 for virtual mode.
Yes: and they used a variation of ASCII that spread the characters out
rather than keeping them in the first 128.
>When the 360 POO was written, the 3101 and 5150 were not products. When
>they wrote it, that's what they intended to happen if the government
>forced ASCII on them. Instead it was market forces that pushed them
>into ASCII. There's a very interesting book in the IBM DP series about
>the history of character codes called _Character Coded Sets_ or
>something like that.
Coded Character Sets: History and Development, from Addison-Wesley's
Systems Programming Series. Amazingly enough, my local public library
has a copy (but they never replaced their copy of David Kahn's _The
Codebreakers_!). However, I saw it first at the local University.
John Savard
Robert Billing
In article <5gl4vs$f...@tor-nn1-hb0.netcom.ca>
sew...@netcom.ca "John Savard" writes:
> But the 'official' standard for ASCII to EBCDIC translation translates
> ASCII ! instead of ASCII | to the EBCDIC vertical bar, and ASCII ^
> instead of ASCII ~ to the logical NOT symbol. Apparently, getting PL/I
> on ASR 33s was more important than the English language.
Does anyome else mourn the passing of the old ascii "up arrow" which
became ^ and the "back arrow" which became _ ? I've had the same ASCII
code chart on a credit card sized card in my wallet for over 20 years,
and, apart from those 2 changes it's still accurate. BTW it has an
advert for Westinghouse W1600 and W1620 terminals on the back.
--
I am Robert Billing, Christian, inventor, traveller, cook and animal
lover, I live near 0:46W 51:22N. "...you're the only thing in this
whole world that's pure & good & right, & wherever you are & wherever
you go there's always...some light." http://www.tnglwood.demon.co.uk/
Charlie Gibbs
In article <5gl4vs$f...@tor-nn1-hb0.netcom.ca> sew...@netcom.ca
(John Savard) writes:
>But the 'official' standard for ASCII to EBCDIC translation translates
>ASCII ! instead of ASCII | to the EBCDIC vertical bar, and ASCII ^
>instead of ASCII ~ to the logical NOT symbol. Apparently, getting PL/I
>on ASR 33s was more important than the English language.
Ah yes, the good old EBCDIC code chart of the week. The vertical
bar used to be hex 4F in EBCDIC, until the revisionists got their
hands on it and moved it to hex 5A where the exclamation mark used
to be. Or was it the previously-unassigned hex 6A so they could
pre-empt hex 4A and 5A for left and right square brackets, which
weren't originally even in the EBCDIC table? I can't remember
where the exclamation mark finally wound up. Maybe back at 4F.
It was all too much for me, and I spent some time reworking printer
translation routines on the mainframe so that if I punched a vertical
bar on a keypunch it would still come out properly on the mainframe
printer, not to mention pass correctly through ASCII translation to
our terminals.
Sheesh.
Charli...@mindlink.bc.ca
Honk if your horn is broken.
bbrey...@aol.com
>Subject: Re: Latest IBM Reference Card?
>From: Eric Jackson <Eric_J...@candle.com>
>Date: Wed, 12 Mar 1997 17:01:31 -0800
>
>John Wilson wrote:
>>
>> In article <332738...@ibmmail.com>,
>> Andy Styles <usfm...@ibmmail.com> wrote:
>> >I've just bought the ESA 390 POp - it cost me just under 12 GBPs I
>...
>> BCDIC, EBCDIC, 029 etc. and then they list ASCII "for completeness", with a
>> footnote saying that no IBM product uses ASCII (funny I thought the 360
>still
>> had an ASCII mode) but if one ever does, they will replace the "^"
>character
>> with a funny weather symbol, or some such bullshit. I love it! Leave no
>> standard unviolated...
>
>The 360 had an ASCII mode, but the only effect was to change the sign
>generated by decimal operations. No IBM operating system ran with that
>bit turned on, and the PSW bit was reused in the 370 for virtual mode.
>
>> AFAIK it was a lie, the 3101 and 5150 use perfectly
>> normal ASCII for printing characters anyway.
>>
>> John Wilson
>> D Bit
>
>When the 360 POO was written, the 3101 and 5150 were not products. When
>they wrote it, that's what they intended to happen if the government
>forced ASCII on them. Instead it was market forces that pushed them
>into ASCII. There's a very interesting book in the IBM DP series about
>the history of character codes called _Character Coded Sets_ or
>something like that.
>
>
The book being referenced is Charles E. Mackenzie, <Coded Character Sets, History and Development>, (The Systems Programming Series), Reading MA, Addison-Wesley, 1980. A follow-up is Jerome J. Andersen, "ASCII, EBCDIC Character Set Committee: Pertinent Character Code Standards", SHARE 69, Session C190 (1981?).
Bill Turner, WB4ALM
> Yes: and they used a variation of ASCII that spread the characters out
> rather than keeping them in the first 128.
>
It wasn't a variation is was a big mistake in reading documentation. The
engineer that read the ASCII doc, made several assumptions and used
those to design the original "communication processors".
Since the translation inside the mainframe form ASCII to EBCDIC used the
TR instruction, the error didn't matter and the equipment was marketed
anyway.
The error?
IBM documents bit 0 as the right hand bit, and ASCII defines bit 0 as
the left hand one. (Or is it the other way around?)
The strange distribution is because the bits are backwards in memory!!!
(and the parity bit COUNTS.) Which is also why IBM likes EVEN parity.
On the other hand, the same translation will handle even/odd/mark/space
parity at the same time.
Now, if they would have standardized the actual code translation
table...
...but that is a time when virtually nobody wanted to use the TN print
chain, cause it degraded the print speed too much.
/s/ Bill Turner, wb4alm
Bill Turner, WB4ALM
>
> > But the 'official' standard for ASCII to EBCDIC translation translates
> > ASCII ! instead of ASCII | to the EBCDIC vertical bar, and ASCII ^
> > instead of ASCII ~ to the logical NOT symbol. Apparently, getting PL/I
> > on ASR 33s was more important than the English language.
>
> became ^ and the "back arrow" which became _ ? I've had the same ASCII
> code chart on a credit card sized card in my wallet for over 20 years,
> and, apart from those 2 changes it's still accurate. BTW it has an
> advert for Westinghouse W1600 and W1620 terminals on the back.
You mean the "to the power of" and "the replaced by" symbols?
How about the "not" symbol, and there was one other strange one, but
I don't remember what it was.
/s/ Bill Turner, wb4alm
Bruce Crabill
I don't know if was done by mistake or not, but yes IBM front ends did
shift the bits out the opposite direction than one would normally do.
I remember trying to configure ASCII lines on a 3705 and it took me a
while to figure out that the characters that you had to code for the
line end characters were bit reversed. Actually, in VM/SP and the 3705,
IBM used MARK parity. One of the mods I did to VM/SP was to go thru all
the hard coded messages in VM and change the parity from MARK to EVEN (the
3705 didn't actually generate or check parity, it just sent 8 bits). This
was before Emacs became popular and made everyone switch from 7 bit with
parity to 8 bits without parity (also was the demise of XON/XOFF flow
control). As for the mainframe, in the 360 there was an ASCII mode bit,
and it controlled how the decimal and edit instructions operated.
Unfortunately for IBM, ASCII hadn't been completely standardized when
they were designing the 360 and in fact, the standards committee ended up
changing ASCII, and breaking IBM's design. In appendix F of my 360 POP
(GA22-6821-8), there is a footnote on the table showing ASCII (which is in
the spread out version mentioned earlier), saying:
"Current activites in committees under the auspices of the United
States of America Standards Institute may result in changes to the
characters and/or structure of the eight-bit representation of USASCII
devised by the Institute. Such changes may cause the eight-bit
representation of USASCII implemented in System/360 (USASCII-8) to be
different from a future USA Standard. Since a difference of this
nature may eventually lead to modification of System/360, it is
recommended that users avoid: (1) operation with PSW bit 12 set to 1,
and (2) the use of any sign codes in decimal data other than those
preferred for EBCDIC."
Bruce
Bill Turner, WB4ALM
Bruce Crabill wrote:
>
>
> I don't know if was done by mistake or not, but yes IBM front ends did
> shift the bits out the opposite direction than one would normally do.
> I remember trying to configure ASCII lines on a 3705 and it took me a
> while to figure out that the characters that you had to code for the
> line end characters were bit reversed. Actually, in VM/SP and the 3705,
> IBM used MARK parity. One of the mods I did to VM/SP was to go thru all
> the hard coded messages in VM and change the parity from MARK to EVEN (the
> 3705 didn't actually generate or check parity, it just sent 8 bits). This
> was before Emacs became popular and made everyone switch from 7 bit with
> parity to 8 bits without parity (also was the demise of XON/XOFF flow
> control).
damn, I did say EVEN didn't I? OOOOOOoooooopppppppps! I knew it was
MARK, because I also made the same VM mods. and for the same reason.
And IBM actually told me in response to several APARs that I was the
only person that wanted the messages in EVEN parity. Well now we know
that there were at least two (thousand or more). <grin>
I always liked that "error" of reversing the bits - everybody thought I
was a Guru or something, because I knew how it really worked.
/s/ Bill Turner, wb4alm
Alan Bowler
In article <AF4FBE75...@0.0.0.0> var...@flash.net (Verne Arase) writes:
>In article <332751...@candle.com>,
>
> >The 360 had an ASCII mode, but the only effect was to change the sign
> >generated by decimal operations. No IBM operating system ran with that
> >bit turned on, and the PSW bit was reused in the 370 for virtual mode.
>
>