Home > Archive > Cobol > December 2007 > diff between cobol1 & cobol3
You are viewing an archived Text-only version of the thread.
To view this thread in it's original format and/or if you want to reply to
this thread please [click here]
| Author |
diff between cobol1 & cobol3
|
|
|
| Can anyone out there point out the major differences in COBOL I and
COBOL III to me? I'm just learning COBOL and I'm kind of 
about
this right now.
| |
|
| In article <45639727-e1dc-42eb-8938-58786661fff8@d27g2000prf.googlegroups.com>,
nishu <nishantchaturvedi@gmail.com> wrote:
>Can anyone out there point out the major differences in COBOL I and
>COBOL III to me? I'm just learning COBOL and I'm kind of
>about
>this right now.
I've been working with COBOL for a few decades now and I think the only
thing I can point out is that I believe one of them exists and one of them
doesn't.
DD
| |
| Rick Smith 2007-12-07, 7:55 am |
|
<docdwarf@panix.com> wrote in message news:fjb6fo$g9h$1@reader1.panix.com...
> In article
<45639727-e1dc-42eb-8938-58786661fff8@d27g2000prf.googlegroups.com>,
> nishu <nishantchaturvedi@gmail.com> wrote:
>
> I've been working with COBOL for a few decades now and I think the only
> thing I can point out is that I believe one of them exists and one of them
> doesn't.
It is not clear to me that either I or III exists; but rather the
designations were contrived to show a relation between II, which
does exist in one popular environment, and the predecessor and
successor implementations for that same environment.
In any case, one "just learning COBOL" might be well-advised
to learn COBOL MCMLXXXV, the intrinsic functions added in
MCMLXXXIX, and the changes introduced in COBOL MMII.
| |
|
| In article <13licj0hkh6b753@corp.supernews.com>,
Rick Smith <ricksmith@mfi.net> wrote:
>
><docdwarf@panix.com> wrote in message news:fjb6fo$g9h$1@reader1.panix.com...
><45639727-e1dc-42eb-8938-58786661fff8@d27g2000prf.googlegroups.com>,
>
>It is not clear to me that either I or III exists; but rather the
>designations were contrived to show a relation between II, which
>does exist in one popular environment, and the predecessor and
>successor implementations for that same environment.
It might be, perhaps, that such a contrivance was attempted... such
foolishments! *Everyone* knows that there was no 'COBOL II'; there *was*
a COBOLII... variants of which appear to include 'cobolii', pronounced as
'koh-bohl-ee' or 'koh-bohl-ai-ai' (the 'ai' being pronounced in a slightly
diphthonic manner reminiscent of exclamations found in Spanish
comic-strips, eg 'El Diablo de Muerte, el bandito grande? Ai! Ai!
Vamenos rapido!').
>In any case, one "just learning COBOL" might be well-advised
>to learn COBOL MCMLXXXV, the intrinsic functions added in
>MCMLXXXIX, and the changes introduced in COBOL MMII.
Durned kids and their newfangled Roman numerals... since when did folks
become 'too good' for scratch-marks on some reindeer-antler?
DD
| |
| Howard Brazee 2007-12-07, 6:56 pm |
| On Fri, 7 Dec 2007 01:40:39 -0800 (PST), nishu
<nishantchaturvedi@gmail.com> wrote:
>Can anyone out there point out the major differences in COBOL I and
>COBOL III to me? I'm just learning COBOL and I'm kind of
>about this right now.
I've never heard of either of them.
But I've been programming CoBOL since 1969, with various computer
systems. Once, when I was working on a VAX, I didn't get a job
because I didn't have "CoBOL II" experience. Later on I found out
that CoBOL II was the same ANSI CoBOL that I was using - except that
name was used by IBM.
CoBOL, like other languages, has evolved some over the years. But
not much. (Unless you are asking about OO CoBOL).
| |
| William M. Klein 2007-12-07, 6:56 pm |
| How and where are you "learning COBOL"?
If you are learning IBM mainframe-specific COBOL, then the chances are that you
are asking for the differences between their products "OS/VS COBOL" and
"Enterprise COBOL".
The former hasn't been supported by IBM for over a decade - and it conformed to
the 1968 and 1974 ANSI Standards. The latter is a currently supported IBM
product and conforms to the 1985 ANSI Standard - with its Amendments. Both have
many IBM extensions (but not the same extensions).
For detailed information on the differences between these two products (along
with the "in-between" product of VS COBOL II), see the IBM migration guide at:
http://publibz.boulder.ibm.com/cgi-...3MG32/CCONTENTS
Again, depending on how you are learning COBOL and whether or not you actually
want to be limited to IBM mainframe COBOL or not, some or none of this
information may be useful.
--
Bill Klein
wmklein <at> ix.netcom.com
"nishu" <nishantchaturvedi@gmail.com> wrote in message
news:45639727-e1dc-42eb-8938-58786661fff8@d27g2000prf.googlegroups.com...
> Can anyone out there point out the major differences in COBOL I and
> COBOL III to me? I'm just learning COBOL and I'm kind of
> about
> this right now.
| |
| Robert 2007-12-07, 6:56 pm |
| On Fri, 7 Dec 2007 01:40:39 -0800 (PST), nishu <nishantchaturvedi@gmail.com> wrote:
>Can anyone out there point out the major differences in COBOL I and
>COBOL III to me? I'm just learning COBOL and I'm kind of
>about
>this right now.
"COBOLIA est omnis divisa in partes tres." Who could forget the famous opening line from
Bellum Cobolicum?
There are significant differences between Roman Cobol I, II and III versus modern Cobol
74, 85 and 02. Roman computers could not represent ZERO, their lowest value was I. This
explains why Cobol subscipts begin with (1) whereas modern languages begin with (0). The
statement PERMOVO MUGIO-PENDO UT FOO (move low-values to foo) filled the area with binary
Is rather than binary zeros. Arithmetic statements producing a value of zero where
considered an error, which had to be handled, for example:
PERMOVO X UT FOO
DEMO X EX FOO
IN AMPLITUDO ERRORIS TRACTARE ERRORIS-TRACTO.
(subtract 10 from foo on size error perform error-handler)
The Roman Standards Committee (workgroup JIV) did not regard ZERO as an integer that could
be stored in memory but rather as the limit of 1/x as x approached infinity. A modern
vestige of this belief is the prohibition against division by zero. x/0 = infinity is a
simple algebraic restatement of 1/infinity = 0. Cobol II was developed during the reign
of Emperor Dodi Schrickerae, who was secretely in league with Interpopulus Res
Aparatus.
Cobol I and II were regarded as Real Cobol. Cobol III, which was widely ignored, was
developed after the fall of Rome by barbarians who worshiped Res Venia (object oriented).
It was later revealed that the barbarians were actually Communists.
Hope this helps.
| |
| William M. Klein 2007-12-07, 6:56 pm |
| However, the Roman COBOL committee is still working on their next revision and
expect to have it out any decade now.
--
Bill Klein
wmklein <at> ix.netcom.com
"Robert" <no@e.mail> wrote in message
news:dokjl31ffeqgd95hbp2bfsq2svcdkrg3lb@
4ax.com...
> On Fri, 7 Dec 2007 01:40:39 -0800 (PST), nishu <nishantchaturvedi@gmail.com>
> wrote:
>
>
> "COBOLIA est omnis divisa in partes tres." Who could forget the famous opening
> line from
> Bellum Cobolicum?
>
> There are significant differences between Roman Cobol I, II and III versus
> modern Cobol
> 74, 85 and 02. Roman computers could not represent ZERO, their lowest value
> was I. This
> explains why Cobol subscipts begin with (1) whereas modern languages begin
> with (0). The
> statement PERMOVO MUGIO-PENDO UT FOO (move low-values to foo) filled the area
> with binary
> Is rather than binary zeros. Arithmetic statements producing a value of zero
> where
> considered an error, which had to be handled, for example:
>
> PERMOVO X UT FOO
> DEMO X EX FOO
> IN AMPLITUDO ERRORIS TRACTARE ERRORIS-TRACTO.
> (subtract 10 from foo on size error perform error-handler)
>
> The Roman Standards Committee (workgroup JIV) did not regard ZERO as an
> integer that could
> be stored in memory but rather as the limit of 1/x as x approached infinity. A
> modern
> vestige of this belief is the prohibition against division by zero. x/0 =
> infinity is a
> simple algebraic restatement of 1/infinity = 0. Cobol II was developed during
> the reign
> of Emperor Dodi Schrickerae, who was secretely in league with Interpopulus
> Res
> Aparatus.
>
> Cobol I and II were regarded as Real Cobol. Cobol III, which was widely
> ignored, was
> developed after the fall of Rome by barbarians who worshiped Res Venia (object
> oriented).
> It was later revealed that the barbarians were actually Communists.
>
> Hope this helps.
| |
|
| In article <dokjl31ffeqgd95hbp2bfsq2svcdkrg3lb@4ax.com>,
Robert <no@e.mail> wrote:
>On Fri, 7 Dec 2007 01:40:39 -0800 (PST), nishu
><nishantchaturvedi@gmail.com> wrote:
>
>
>"COBOLIA est omnis divisa in partes tres." Who could forget the famous
>opening line from
>Bellum Cobolicum?
Oh, I *cannot* resist...
.... that kind of false quoting, Mr Wagner, shows 300% more 'brazen
boldness coupled with impudent assurance and insolence' than should be
permitted to remain joined together... yes, it has enough gall to be
divided into three parts.
DD
| |
| Robert 2007-12-07, 9:56 pm |
| On Sat, 8 Dec 2007 01:09:14 +0000 (UTC), docdwarf@panix.com () wrote:
>In article <dokjl31ffeqgd95hbp2bfsq2svcdkrg3lb@4ax.com>,
>Robert <no@e.mail> wrote:
>
>Oh, I *cannot* resist...
>
>... that kind of false quoting, Mr Wagner, shows 300% more 'brazen
>boldness coupled with impudent assurance and insolence' than should be
>permitted to remain joined together... yes, it has enough gall to be
>divided into three parts.
Didn't you mean insouciance (marked by blythe unconcern), "an elegantly insouciant
manner?"
| |
|
| In article <qjvjl3t8ag0gpjqml2al8tn3jb149rapve@4ax.com>,
Robert <no@e.mail> wrote:
>On Sat, 8 Dec 2007 01:09:14 +0000 (UTC), docdwarf@panix.com () wrote:
>
>
>Didn't you mean insouciance (marked by blythe unconcern), "an elegantly
>insouciant
>manner?"
Not in the least...you see, while it might have been accompanied - strange
as some migh find it! - by a modicum of good sense there were sufficient
quantities of neither to degrade the human condition... so for this
instance good sense, insouciance did *not* cripple mankind.
(note - the puns, literary and social references here might be a bit
outside of the purview of folks unfamiliar with certain things... feel
free to ask; ignorance can be cured)
DD
| |
| Robin Lee 2007-12-08, 6:55 pm |
| Perhaps the question is regarding level 1 or level 2.
Several decades ago when first learning COBOL 74 on
a particular platform, there was some mention in the
manual that theirs included all Level I features and
some Level II. It appeared most of what they considered
Level II had to do with their implementation of ISAM.
| |
| Alistair 2007-12-08, 6:55 pm |
| On 8 Dec, 11:09, docdw...@panix.com () wrote:
> In article <qjvjl3t8ag0gpjqml2al8tn3jb149ra...@4ax.com>,
>
>
>
>
>
> Robert <n...@e.mail> wrote:
>
>
>
>
>
>
>
> Not in the least...you see, while it might have been accompanied - strange
> as some migh find it! - by a modicum of good sense there were sufficient
> quantities of neither to degrade the human condition... so for this
> instance good sense, insouciance did *not* cripple mankind.
>
> (note - the puns, literary and social references here might be a bit
> outside of the purview of folks unfamiliar with certain things... feel
> free to ask; ignorance can be cured)
>
> DD- Hide quoted text -
>
Sometimes it is better to remain silent and to be considered ignorant
than to open one's mouth and to remove all doubt.
| |
| Alistair 2007-12-08, 6:55 pm |
| On 7 Dec, 09:40, nishu <nishantchaturv...@gmail.com> wrote:
> Can anyone out there point out the major differences in COBOL I and
> COBOL III to me? I'm just learning COBOL and I'm kind of
> about
> this right now.
Cobol I is Ansi 74 compliant (-ish) and Cobol II is Ansi 85 compliant
(-ish). The only reference to Cobol III that I have found is in
reference to deducing 666-code from existing cobol programs.
Cobol I and II are IBM variants of Cobol. To my knowledge, neither is
supported any more, with Cobol for MVS and another Cobol (? don't
recall the name) being the replacements.
Cobol 85 (the Cobol II standard) uses such commands as EVALUATE
(spit!) for those who don't understand complex IF structires. I can
not remember what else it does but you could look up the standards
documents. It also adds END-IF or IF-END commands and replaces NEXT
SENTENCE (for the criminally minded Coboller) with CONTINUE
(unapprehended).
Be advised that Cobol for MVS does stupid things like changing the
date formats (which might be over-ridable with compiler directives but
wasn't where I worked for the year 2000 project).
| |
| William M. Klein 2007-12-08, 6:55 pm |
| If this is a serious post, then let me explain.
For the '68 and '74 and '85 Standards (but not the '02 Standard), the Standard
was divided into "modules" (e.g. Nucleus, Sequential I/O, Sort-Merge, Debug,
etc). Some, but not all of these modules had two defined "levels". For
example, "COPY" was in Level 1 of the "text manipulation module" (I can't
remember its exact name) while "COPY/REPLACING" was in level 2.
There were NIST (FIPS) tests for "flagging" of which module certain syntax was
in. All of this was pretty obscure and not important to most programmers, but
did impact what level of "conformance" an implementer could claim.
(It was related to - but not quite identical to - claiming "high" "intermediate"
or "low" level of conformance for the entire compiler).
--
Bill Klein
wmklein <at> ix.netcom.com
"Robin Lee" <robinlee@news.com> wrote in message
news:Ze6dnSTvYZ84NsfanZ2dnUVZ_gydnZ2d@gi
ganews.com...
> Perhaps the question is regarding level 1 or level 2.
> Several decades ago when first learning COBOL 74 on
> a particular platform, there was some mention in the
> manual that theirs included all Level I features and
> some Level II. It appeared most of what they considered
> Level II had to do with their implementation of ISAM.
| |
| William M. Klein 2007-12-08, 6:55 pm |
| A bit of clarification,
1) IBM used the name "VS COBOL II" (not "COBOL II). They NEVER used the names
"COBOL I" (or COBOL 1) or "COBOL III (or "COBOL 3"). CICS had translator
options called "COBOL2" and "COBOL3" and IBM did sell a "COBOL/2" product.
2) OS/VS COBOL (what may have been considered "COBOL 1" - although it actually
was a follow-on product to ANS COBOL V4) fully supported the '68 and '74 ANSI
Standards. (The LANGLVL compiler option impacted this).
3) VS COBOL II Releases 1, 1.0, and 2 were ONLY '74 Standard compilers. With VS
COBOL II, R3 (thru R4), the compiler was PRIMARILY a (fully conforming) '85
Standard compiler. Even with those releases it MOSTLY supported the '74
Standard via the CMPR2 compiler option.
4) All IBM mainframe compilers that require (not allow) LE as their run-time,
i.e. SAA/370 COBOL, COBOL for MVS & VM, COBOL for z/OS, and the currently
supported Enterprise COBOL, support the '85 Standard *plus* the Intrinsic
Function module. The current (Enterprise COBOL) no longer supports the CMPR2
compiler option to provide '74 Standard support.
5) Although there are *Minor* syntax compatibility issues in migrating older
COBOL to currently supported (IBM) COBOL, I am not aware of any "date format"
issues. However, the Migration Guide that I pointed to in my original note does
give details on ALL the differences between the current Enterprise COBOL and
previous (VS COBOL II and OS/VS COBOL) compilers.
6) "next sentence" is fully supported in the '85 Standard and is NOT equivalent
to "continue" - except in relatively simplistic examples - and examples without
End-xxxx structures. (And before anyone asks, IBM does support NEXT SENTENCE
with END-IF - although the Standard doesn't for most cases.)
--
Bill Klein
wmklein <at> ix.netcom.com
"Alistair" <alistair@ld50macca.demon.co.uk> wrote in message
news:7c7e163f-835e-4a38-8fbb-c52c71d952eb@s19g2000prg.googlegroups.com...
> On 7 Dec, 09:40, nishu <nishantchaturv...@gmail.com> wrote:
>
> Cobol I is Ansi 74 compliant (-ish) and Cobol II is Ansi 85 compliant
> (-ish). The only reference to Cobol III that I have found is in
> reference to deducing 666-code from existing cobol programs.
>
> Cobol I and II are IBM variants of Cobol. To my knowledge, neither is
> supported any more, with Cobol for MVS and another Cobol (? don't
> recall the name) being the replacements.
>
> Cobol 85 (the Cobol II standard) uses such commands as EVALUATE
> (spit!) for those who don't understand complex IF structires. I can
> not remember what else it does but you could look up the standards
> documents. It also adds END-IF or IF-END commands and replaces NEXT
> SENTENCE (for the criminally minded Coboller) with CONTINUE
> (unapprehended).
>
> Be advised that Cobol for MVS does stupid things like changing the
> date formats (which might be over-ridable with compiler directives but
> wasn't where I worked for the year 2000 project).
| |
| Kelly Bert Manning 2007-12-09, 6:55 pm |
| "William M. Klein" (wmklein@nospam.netcom.com) writes:
> A bit of clarification,
>
> 1) IBM used the name "VS COBOL II" (not "COBOL II). They NEVER used the names
> "COBOL I" (or COBOL 1) or "COBOL III (or "COBOL 3"). CICS had translator
> options called "COBOL2" and "COBOL3" and IBM did sell a "COBOL/2" product.
>
> 2) OS/VS COBOL (what may have been considered "COBOL 1" - although it actually
> was a follow-on product to ANS COBOL V4) fully supported the '68 and '74 ANSI
> Standards. (The LANGLVL compiler option impacted this).
I came to pretty much the same conclusion in 1979 after local IBMers told
me that neither ANS or OS/VS COBOL corresponded to the '68 or '74 standards.
My recollection is that the paper documentation was pretty skimpy on that
issue. I concluded OS/VS COBOL was '74 by noting that it accepted code ported
from our Honeywell with fewer objections than the ANS compiler had on older
machines.
There was an older, or at least non-optimizing compiler on one VS2 machine
I had to use around the same time. The performance difference compared
to the optimizing compiler on another model 145 in a VS2 data center 10 blocks
away was quite obvious and repeatable.
I also used a DOS COBOL compiler on a model 135 across the street from one
of the 145s. Can't recall what standard it was.
| |
| Pete Dashwood 2007-12-09, 9:56 pm |
|
"Robert" <no@e.mail> wrote in message
news:dokjl31ffeqgd95hbp2bfsq2svcdkrg3lb@
4ax.com...
> On Fri, 7 Dec 2007 01:40:39 -0800 (PST), nishu
> <nishantchaturvedi@gmail.com> wrote:
>
>
> "COBOLIA est omnis divisa in partes tres." Who could forget the famous
> opening line from
> Bellum Cobolicum?
>
> There are significant differences between Roman Cobol I, II and III versus
> modern Cobol
> 74, 85 and 02. Roman computers could not represent ZERO, their lowest
> value was I. This
> explains why Cobol subscipts begin with (1) whereas modern languages begin
> with (0). The
> statement PERMOVO MUGIO-PENDO UT FOO (move low-values to foo) filled the
> area with binary
> Is rather than binary zeros. Arithmetic statements producing a value of
> zero where
> considered an error, which had to be handled, for example:
>
> PERMOVO X UT FOO
> DEMO X EX FOO
> IN AMPLITUDO ERRORIS TRACTARE ERRORIS-TRACTO.
> (subtract 10 from foo on size error perform error-handler)
>
> The Roman Standards Committee (workgroup JIV) did not regard ZERO as an
> integer that could
> be stored in memory but rather as the limit of 1/x as x approached
> infinity. A modern
> vestige of this belief is the prohibition against division by zero. x/0 =
> infinity is a
> simple algebraic restatement of 1/infinity = 0. Cobol II was developed
> during the reign
> of Emperor Dodi Schrickerae, who was secretely in league with
> Interpopulus Res
> Aparatus.
>
> Cobol I and II were regarded as Real Cobol. Cobol III, which was widely
> ignored, was
> developed after the fall of Rome by barbarians who worshiped Res Venia
> (object oriented).
> It was later revealed that the barbarians were actually Communists.
>
> Hope this helps.
ROFL!
Great stuff, Robert...
Pete.
| |
| Pete Dashwood 2007-12-09, 9:56 pm |
|
"William M. Klein" <wmklein@nospam.netcom.com> wrote in message
news:Bul6j.23573$nc1.881@fe02.news.easynews.com...
> However, the Roman COBOL committee is still working on their next revision
> and expect to have it out any decade now.
>
LOL!
Bit too close to the bone to be REALLY funny :-)
Pete.
--
"I used to write COBOL...now I can do anything."
| |
| Howard Brazee 2007-12-10, 6:55 pm |
| On Sat, 8 Dec 2007 01:09:14 +0000 (UTC), docdwarf@panix.com () wrote:
>... that kind of false quoting, Mr Wagner, shows 300% more 'brazen
>boldness coupled with impudent assurance and insolence' than should be
>permitted to remain joined together... yes, it has enough gall to be
>divided into three parts.
Good, but not Walt Kelly good.
| |
|
| In article <lamql3tpuhbasdgeah2kig47ii3netnegf@4ax.com>,
Howard Brazee <howard@brazee.net> wrote:
>On Sat, 8 Dec 2007 01:09:14 +0000 (UTC), docdwarf@panix.com () wrote:
>
>
>Good, but not Walt Kelly good.
He does his job, I try to do mine.
DD
| |
| Charles Hottel 2007-12-10, 6:55 pm |
|
"Pete Dashwood" <dashwood@removethis.enternet.co.nz> wrote in message
news:5s3hiaF16s96hU1@mid.individual.net...
>
>
> "William M. Klein" <wmklein@nospam.netcom.com> wrote in message
> news:Bul6j.23573$nc1.881@fe02.news.easynews.com...
> LOL!
>
> Bit too close to the bone to be REALLY funny :-)
>
> Pete.
> --
> "I used to write COBOL...now I can do anything."
>
>
Probably the singularity will get here first.
| |
| Robert 2007-12-11, 9:56 pm |
| On Sat, 8 Dec 2007 07:14:23 -0800 (PST), Alistair <alistair@ld50macca.demon.co.uk> wrote:
>Be advised that Cobol for MVS does stupid things like changing the
>date formats (which might be over-ridable with compiler directives but
>wasn't where I worked for the year 2000 project).
Except intrinsic functions and ACCEPT .. FROM DATE, Cobol did not have a date format until
IBM introduced the DATE FORMAT clause in the late '90s. IBM did not change intrinsic
functions nor ACCEPT except to add an 02-style YYYYMMDD option.
What are you referring to?
| |
| William M. Klein 2007-12-11, 9:56 pm |
| I also questioned this in one of my replies. The thing that they MIGHT have
been thinking of was the IBM extension of a current-date special register. This
is no longer available. See (for example)
http://publibz.boulder.ibm.com/cgi-.../igy3mg32/4.1.5
and look for the topic,
"CURRENT-DATE special register"
The only other thing that I could think of that he might have been thinking
about is the INTDATE compiler option. See:
http://publibz.boulder.ibm.com/cgi-...IGY3PG32/2.4.25
which deals with what COBOL vs LE thinks of as "day 1" for Lilian (or
pseudo-Lilian) dates.
--
Bill Klein
wmklein <at> ix.netcom.com
"Robert" <no@e.mail> wrote in message
news:lscul35p5nef2qjrs2de41vipuv82aifor@
4ax.com...
> On Sat, 8 Dec 2007 07:14:23 -0800 (PST), Alistair
> <alistair@ld50macca.demon.co.uk> wrote:
>
>
> Except intrinsic functions and ACCEPT .. FROM DATE, Cobol did not have a date
> format until
> IBM introduced the DATE FORMAT clause in the late '90s. IBM did not change
> intrinsic
> functions nor ACCEPT except to add an 02-style YYYYMMDD option.
>
> What are you referring to?
| |
| Robert 2007-12-12, 3:55 am |
| On Wed, 12 Dec 2007 01:48:21 GMT, "William M. Klein" <wmklein@nospam.netcom.com> wrote:
>I also questioned this in one of my replies. The thing that they MIGHT have
>been thinking of was the IBM extension of a current-date special register. This
>is no longer available. See (for example)
>
> http://publibz.boulder.ibm.com/cgi-.../igy3mg32/4.1.5
>
>and look for the topic,
> "CURRENT-DATE special register"
Conversion is easier than the IBM manual shows.
01 new-DATE pic 9(6).
01 old-CURRENT-DATE pic 99/99/99.
accept new-DATE from DATE
compute old-CURRENT-DATE = new-DATE * 100.0001
>The only other thing that I could think of that he might have been thinking
>about is the INTDATE compiler option. See:
> http://publibz.boulder.ibm.com/cgi-...IGY3PG32/2.4.25
>
>which deals with what COBOL vs LE thinks of as "day 1" for Lilian (or
>pseudo-Lilian) dates.
IBM got it right. The Gregorian calendar did begin on 10/15/1582, in Italy and Catholic
places. England and its (then) Dominions switched to the Gregorian calendar on 9/14/1752.
Nothing notable date-wise happened in 1601. Why is the Cobol floor 01/01/1601?
| |
| William M. Klein 2007-12-12, 3:55 am |
| "Robert" <no@e.mail> wrote in message
news:iglul39116trio9td6bnuj81gc2r60nkhd@
4ax.com...
> On Wed, 12 Dec 2007 01:48:21 GMT, "William M. Klein"
> <wmklein@nospam.netcom.com> wrote:
<snip>
>
> IBM got it right. The Gregorian calendar did begin on 10/15/1582, in Italy and
> Catholic
> places. England and its (then) Dominions switched to the Gregorian calendar
> on 9/14/1752.
> Nothing notable date-wise happened in 1601. Why is the Cobol floor 01/01/1601?
The COBOL definition exists so that when you use the MOD and 7 intrinsic
function you will get the same results for day-of-w - other than "0" vs "7".
As VERY FEW COBOL programs need to deal with dates between 1582 and 1601, I
think they thought this difference was less important that MOD / 7 results.
P.S. I think there may be some Library programs that need to deal with those
pre-1601 dates, but probably not as many as COBOL programs needing to deal with
Unicode Klingon data <G>.
--
Bill Klein
wmklein <at> ix.netcom.com
| |
| Robert 2007-12-12, 3:55 am |
| On Wed, 12 Dec 2007 05:30:34 GMT, "William M. Klein" <wmklein@nospam.netcom.com> wrote:
>"Robert" <no@e.mail> wrote in message
> news:iglul39116trio9td6bnuj81gc2r60nkhd@
4ax.com...
><snip>
>
>The COBOL definition exists so that when you use the MOD and 7 intrinsic
>function you will get the same results for day-of-w - other than "0" vs "7".
>As VERY FEW COBOL programs need to deal with dates between 1582 and 1601, I
>think they thought this difference was less important that MOD / 7 results.
Lilian date yields exactly the same day of w. 10/15/1582 was on a Monday.
Both return 0=Sunday, 1=Monday ... 6=Saturday.
| |
| William M. Klein 2007-12-12, 3:55 am |
| I wasn't there, but according to
http://en.wikipedia.org/wiki/1582
"However, the next day became Friday, October 15 ..."
Similarly,
http://robitaille.wordpress.com/200...september-1752/
"When that was done, they had to get rid of 10 days to re-adjust the yearly
cycle. So for the Church, the last day of the Julian calendar was Thursday
October 4 1582, which was followed the next morning by the first day of the
Gregorian calendar, Friday October 15 1582."
--
Bill Klein
wmklein <at> ix.netcom.com
"Robert" <no@e.mail> wrote in message
news:eh0vl3tn87qu67vu5l4n9sjlrnq2r19tk6@
4ax.com...
> On Wed, 12 Dec 2007 05:30:34 GMT, "William M. Klein"
> <wmklein@nospam.netcom.com> wrote:
>
>
> Lilian date yields exactly the same day of w. 10/15/1582 was on a Monday.
> Both return 0=Sunday, 1=Monday ... 6=Saturday.
| |
| Robert 2007-12-12, 6:55 pm |
| On Wed, 12 Dec 2007 07:37:56 GMT, "William M. Klein" <wmklein@nospam.netcom.com> wrote:
>I wasn't there, but according to
> http://en.wikipedia.org/wiki/1582
>
>"However, the next day became Friday, October 15 ..."
>
>Similarly,
> http://robitaille.wordpress.com/200...september-1752/
>
>"When that was done, they had to get rid of 10 days to re-adjust the yearly
>cycle. So for the Church, the last day of the Julian calendar was Thursday
>October 4 1582, which was followed the next morning by the first day of the
>Gregorian calendar, Friday October 15 1582."
You are correct. I was going by this calendar, which shows October 15 1582 to be Monday:
http://www.genuki.org.uk/big/easter/B.html?1582
The explanation is that it's an ENGLISH calendar, therefore it's Julian.
>"Robert" <no@e.mail> wrote in message
> news:eh0vl3tn87qu67vu5l4n9sjlrnq2r19tk6@
4ax.com...
>
| |
| Howard Brazee 2007-12-12, 6:55 pm |
| On Tue, 11 Dec 2007 23:20:07 -0600, Robert <no@e.mail> wrote:
>Nothing notable date-wise happened in 1601.
When was DocDwarf's birthdate?
| |
|
| In article <0950m3pfm976brjsdtdje7o79t8gmkkcet@4ax.com>,
Howard Brazee <howard@brazee.net> wrote:
>On Tue, 11 Dec 2007 23:20:07 -0600, Robert <no@e.mail> wrote:
>
>
>When was DocDwarf's birthdate?
That, most likely, depends on which calendar one uses.
DD
| |
| tlmfru 2007-12-12, 6:55 pm |
| If you're into genealogy and family history, you have to be very aware of
this. The fifteen days isn't all that much of a problem but the three
months were! A child might be buried "six" months after his birth date,
aged nine months, because of this.
I've read that there was an awful kerfuffle over the missing 10 days: from
people thinking they'd lost ten days of their life, to workers being paid
for only twenty days but being billed for thirty, to churches worried about
losing their full tithes ... lots of turmoil.
PL
William M. Klein <wmklein@nospam.netcom.com> wrote in message
news:odM7j.89017$ys5.50005@fe03.news.easynews.com...
> I wasn't there, but according to
> http://en.wikipedia.org/wiki/1582
>
> "However, the next day became Friday, October 15 ..."
>
> Similarly,
>
http://robitaille.wordpress.com/200...september-1752/
>
> "When that was done, they had to get rid of 10 days to re-adjust the
yearly
> cycle. So for the Church, the last day of the Julian calendar was Thursday
> October 4 1582, which was followed the next morning by the first day of
the
> Gregorian calendar, Friday October 15 1582."
>
> --
> Bill Klein
> wmklein <at> ix.netcom.com
> "Robert" <no@e.mail> wrote in message
> news:eh0vl3tn87qu67vu5l4n9sjlrnq2r19tk6@
4ax.com...
thinking[color=darkred]
http://publibz.boulder.ibm.com/cgi-...IGY3PG32/2.4.25[color=darkred]
Italy[color=darkred]
calendar[color=darkred]
vs[color=darkred]
1601, I[color=darkred]
results.[color=darkred]
Monday.[color=darkred]
>
>
| |
| William M. Klein 2007-12-12, 6:55 pm |
| When the ANSI intrinsic functions first came out (were approved), I actually
submitted an interpretation request concerning some of the Julian/Gregorian
calendar issues. Specifically, as some of the functions have "offset from GMT"
values, I asked whether the "date-of-integer" function was or was not sensitive
to where the application was run. Specifically, as the US was a part of the
British "domain" (was it considered an Empire in those days) between 1601 and
the time the Gregorian Calendar was adopted, which calendar's "date" was
returned if a program was run in the US???
The answer (as I expected) was that as far as COBOL was concerned, dates are
always processed as if the Gregorian calendar were in effect for all days in all
places that a COBOL application was compiled and run. (Actually, more of an
issue for ISO COBOL - with Russia being so late in adopting it).
--
Bill Klein
wmklein <at> ix.netcom.com
"tlmfru" <lacey@mts.net> wrote in message
news:8WU7j.11454$Fa7.3120@newsfe17.lga...
> If you're into genealogy and family history, you have to be very aware of
> this. The fifteen days isn't all that much of a problem but the three
> months were! A child might be buried "six" months after his birth date,
> aged nine months, because of this.
>
> I've read that there was an awful kerfuffle over the missing 10 days: from
> people thinking they'd lost ten days of their life, to workers being paid
> for only twenty days but being billed for thirty, to churches worried about
> losing their full tithes ... lots of turmoil.
>
> PL
> William M. Klein <wmklein@nospam.netcom.com> wrote in message
> news:odM7j.89017$ys5.50005@fe03.news.easynews.com...
> http://robitaille.wordpress.com/200...september-1752/
> yearly
> the
> thinking
> http://publibz.boulder.ibm.com/cgi-...IGY3PG32/2.4.25
> Italy
> calendar
> vs
> 1601, I
> results.
> Monday.
>
>
>
>
| |
| Alistair 2007-12-17, 6:56 pm |
| On 12 Dec, 01:12, Robert <n...@e.mail> wrote:
> On Sat, 8 Dec 2007 07:14:23 -0800 (PST), Alistair <alist...@ld50macca.demon.co.uk> wrote:
>
> Except intrinsic functions and ACCEPT .. FROM DATE, Cobol did not have a date format until
> IBM introduced the DATE FORMAT clause in the late '90s. IBM did not change intrinsic
> functions nor ACCEPT except to add an 02-style YYYYMMDD option.
>
> What are you referring to?
Unfortunately I can not recall which format the date came in in. What
I do recall is that in testing I spotted that the date printed in one
(amended) program was in the wrong format and that was after the
program (and all others) had been changed to cater for the format
being passed being at variance to that which we had previously used. I
have wondered whether the format can be over-ridden by compiler
options but that is really something that I no longer have any
interest in.
| |
| Alistair 2007-12-17, 6:56 pm |
| On 12 Dec, 01:48, "William M. Klein" <wmkl...@nospam.netcom.com>
wrote:
> I also questioned this in one of my replies. The thing that they MIGHT have
> been thinking of was the IBM extension of a current-date special register. This
> is no longer available. See (for example)
>
> http://publibz.boulder.ibm.com/cgi-...S/igy3mg32/4...
>
> and look for the topic,
> "CURRENT-DATE special register"
>
> The only other thing that I could think of that he might have been thinking
> about is the INTDATE compiler option. See:
> http://publibz.boulder.ibm.com/cgi-...S/IGY3PG32/2...
>
> which deals with what COBOL vs LE thinks of as "day 1" for Lilian (or
> pseudo-Lilian) dates.
>
> --
> Bill Klein
> wmklein <at> ix.netcom.com"Robert" <n...@e.mail> wrote in message
>
> news:lscul35p5nef2qjrs2de41vipuv82aifor@
4ax.com...
>
>
>
>
>
>
>
> - Show quoted text -
It would have been the ACCEPT function that we used. See my earlier
reply. However, I can not recall which format we were expecteing nor
which format we received but we did change every program which used
dates in order to accomodate the Cobol for MVS compiler which did not
return the date in the format we expected. I don't think this has
anything to do with the US habit of usiong dates in the YYYYDDMM
format c.f. the UK where we use YYYYMMDD.
| |
| Alistair 2007-12-17, 6:56 pm |
| On 12 Dec, 01:12, Robert <n...@e.mail> wrote:
> On Sat, 8 Dec 2007 07:14:23 -0800 (PST), Alistair <alist...@ld50macca.demon.co.uk> wrote:
>
> Except intrinsic functions and ACCEPT .. FROM DATE, Cobol did not have a date format until
> IBM introduced the DATE FORMAT clause in the late '90s. IBM did not change intrinsic
> functions nor ACCEPT except to add an 02-style YYYYMMDD option.
>
> What are you referring to?
I have the feeling that my second reply to this post was meant to be
to WMK. Sorry folks, for any confusion.
| |
| Alistair 2007-12-17, 6:56 pm |
| On 12 Dec, 05:20, Robert <n...@e.mail> wrote:
> On Wed, 12 Dec 2007 01:48:21 GMT, "William M. Klein" <wmkl...@nospam.netcom.com> wrote:
>
>
>
>
> Conversion is easier than the IBM manual shows.
>
> 01 new-DATE pic 9(6).
> 01 old-CURRENT-DATE pic 99/99/99.
>
> accept new-DATE from DATE
> compute old-CURRENT-DATE = new-DATE * 100.0001
>
>
>
> IBM got it right. The Gregorian calendar did begin on 10/15/1582, in Italy and Catholic
> places. England and its (then) Dominions switched to the Gregorian calendar on 9/14/1752.
> Nothing notable date-wise happened in 1601. Why is the Cobol floor 01/01/1601?
Aren't you guys lucky that you were still part of such a forward-
looking empire in 1752 and joined the Gregorian world. Otherwise you'd
be celebrating Christmas on the 6th of January. ;-)
| |
| Alistair 2007-12-17, 6:56 pm |
| On 12 Dec, 05:30, "William M. Klein" <wmkl...@nospam.netcom.com>
wrote:
> "Robert" <n...@e.mail> wrote in message
>
> news:iglul39116trio9td6bnuj81gc2r60nkhd@
4ax.com...
>
> <snip>
>
>
>
> The COBOL definition exists so that when you use the MOD and 7 intrinsic
> function you will get the same results for day-of-w - other than "0" vs "7".
> As VERY FEW COBOL programs need to deal with dates between 1582 and 1601, I
> think they thought this difference was less important that MOD / 7 results.
>
> P.S. I think there may be some Library programs that need to deal with those
> pre-1601 dates, but probably not as many as COBOL programs needing to deal with
> Unicode Klingon data <G>.
>
> --
> Bill Klein
> wmklein <at> ix.netcom.com
Are you suggesting that we should incorporate Klingon Unicode into the
next Cobol standard? ;-)
| |
| Alistair 2007-12-17, 6:56 pm |
| On 12 Dec, 17:31, "tlmfru" <la...@mts.net> wrote:
> If you're into genealogy and family history, you have to be very aware of
> this. The fifteen days isn't all that much of a problem but the three
> months were! A child might be buried "six" months after his birth date,
> aged nine months, because of this.
>
> I've read that there was an awful kerfuffle over the missing 10 days: from
> people thinking they'd lost ten days of their life, to workers being paid
> for only twenty days but being billed for thirty, to churches worried about
> losing their full tithes ... lots of turmoil.
>
No problem with my family. I can trace us back to about 1650 but not
much earlier.
| |
| Alistair 2007-12-17, 6:56 pm |
| On 12 Dec, 20:38, "William M. Klein" <wmkl...@nospam.netcom.com>
wrote:
> Specifically, as the US was a part of the
> British "domain" (was it considered an Empire in those days)
Absolutely so.
> between 1601 and
> the time the Gregorian Calendar was adopted, which calendar's "date" was
> returned if a program was run in the US???
>
> The answer (as I expected) was that as far as COBOL was concerned, dates are
> always processed as if the Gregorian calendar were in effect for all days in all
> places that a COBOL application was compiled and run. (Actually, more of an
> issue for ISO COBOL - with Russia being so late in adopting it).
>
So, Gregorian dates are applied by Cobol applications back before 1601
regardless of whether the calendar applied at the time was Julian or
other?
| |
| Howard Brazee 2007-12-17, 6:56 pm |
| On Mon, 17 Dec 2007 13:48:54 -0800 (PST), Alistair
<alistair@ld50macca.demon.co.uk> wrote:
>No problem with my family. I can trace us back to about 1650 but not
>much earlier.
Of course, go back 10 generations and find somewhere around 1024
different family threads. What percentage of these can you trace?
| |
| Howard Brazee 2007-12-17, 6:56 pm |
| On Mon, 17 Dec 2007 13:45:30 -0800 (PST), Alistair
<alistair@ld50macca.demon.co.uk> wrote:
>Aren't you guys lucky that you were still part of such a forward-
>looking empire in 1752 and joined the Gregorian world. Otherwise you'd
>be celebrating Christmas on the 6th of January. ;-)
I work with an ex-Lithuanian someone who celebrates it the 6th of
January. And December 25th.
| |
| Howard Brazee 2007-12-17, 6:56 pm |
| On Mon, 17 Dec 2007 14:59:19 -0700, Howard Brazee <howard@brazee.net>
wrote:
>
>I work with an ex-Lithuanian someone who celebrates it the 6th of
>January. And December 25th.
Ukrainian. I don't know where that came from. Maybe Captain Ramos.
| |
| Robert 2007-12-18, 3:55 am |
| On Mon, 17 Dec 2007 13:52:03 -0800 (PST), Alistair <alistair@ld50macca.demon.co.uk> wrote:
>On 12 Dec, 20:38, "William M. Klein" <wmkl...@nospam.netcom.com>
>wrote:
>
>Absolutely so.
>
>
>
>So, Gregorian dates are applied by Cobol applications back before 1601
>regardless of whether the calendar applied at the time was Julian or
>other?
Our biggest nightmare was in -46, when the Julian calendar was adopted. In order to get
his bright idea in sync with the seasons, Julius, the guy in the corner office, added
three months, creating a year with 445 days. We had to rewrite ALL our date routines.
BTW, Julius hated the yyddd format and forbade its use within the Pax Romana empire.
It was introduced during the Dark Ages by some barbarians, who mockingly called it the
"julian date". The joke was lost on later programmers, who mistakenly believed Julius had
something to do with it. Come on folks, we weren't THAT backward.
Just before Y1K, we lobbied the Cobol Standards committee to increase the year from three
digits to four. Unexpectedly, they came up with a TWO digit year. Their only rationale
involved job security every century. I think the guy in charge worked for Tata.
During the 170 years of two calendars, date conversion utilities were a thriving industry.
Our friends in Russia and Greece kept the scam going into the twentieth century. All
that's left are a few programs maintained by monks in the Orthodox Church, who still
number their paragraphs because it makes them easy to find. Their source code is chiseled
in stone tablets (from whence we get the practice of never deleting old code; they have a
chisel shaped like an asterisk). On the edge of each tablet is chiseled the 'topic' or
range of paragraph numbers. For instance M contains paragraphs M through MCMXCIX, MM
contains MM through MMCMXCIX, etc. They did not use names after the numbers to reduce
development (chiseling) time.
| |
| Robert 2007-12-18, 3:55 am |
| On Mon, 17 Dec 2007 13:46:47 -0800 (PST), Alistair <alistair@ld50macca.demon.co.uk> wrote:
>On 12 Dec, 05:30, "William M. Klein" <wmkl...@nospam.netcom.com>
>wrote:
>
>Are you suggesting that we should incorporate Klingon Unicode into the
>next Cobol standard? ;-)
ISO/IEC 10646 includes Klingon Unicode.
http://www.evertype.com/standards/csur/klingon.html
I don't believe it includes EBCDIC. They rejected it as 'too wierd'.
| |
| Howard Brazee 2007-12-19, 6:56 pm |
| On Tue, 18 Dec 2007 00:32:02 -0600, Robert <no@e.mail> wrote:
>During the 170 years of two calendars, date conversion utilities were a thriving industry.
>Our friends in Russia and Greece kept the scam going into the twentieth century. All
>that's left are a few programs maintained by monks in the Orthodox Church, who still
>number their paragraphs because it makes them easy to find. Their source code is chiseled
>in stone tablets (from whence we get the practice of never deleting old code; they have a
>chisel shaped like an asterisk). On the edge of each tablet is chiseled the 'topic' or
>range of paragraph numbers. For instance M contains paragraphs M through MCMXCIX, MM
>contains MM through MMCMXCIX, etc. They did not use names after the numbers to reduce
>development (chiseling) time.
The scam continues with time zones, and especially daylight savings
time.
| |
| Alistair 2007-12-19, 6:56 pm |
| On 17 Dec, 21:57, Howard Brazee <how...@brazee.net> wrote:
> On Mon, 17 Dec 2007 13:48:54 -0800 (PST), Alistair
>
> <alist...@ld50macca.demon.co.uk> wrote:
>
> Of course, go back 10 generations and find somewhere around 1024
> different family threads. What percentage of these can you trace?
2 threads.
| |
|
|
|
|
|