tiny-cobol-users

Re: [Tiny-cobol-users] Tiny-cobol-users Digest, Vol 7, Issue 1

[John R. C. <jo...@we...>]

On Saturday 23 December 2006 20:49, wsi...@op... wrote:
> Hi John,As I read this list, your request for help came to my attention.Not
> that I could help, but perhaps you could help me.My long time connection to
> CODASYL keeps me interested in all things COBOL.And DATABASE/ This practice
> makes an old man feel like he is still envolved in some way with hismajor
> work assignments prior to retirement almost 25 years ago.There was ONE
> sentence that is responsible for this enquiry. The one which saysyou have
> CC and TC both enstalled and running on your system. I think I rememberyou
> have published some COBOL documents (perhase books) on the subject.So, my
> first interest is what is your system. If I asked for and you agreed to
> provideme with all it would take to have the same ability, would my system
> be large enough.I currently run a HP Pavilion q1520y which backs up the C
> drive with a D drive I can notaccess.There is an A drive, A zip drive, and
> a HPDVD drive as well as two printers. The screenis a large flat collor
> held over from my former Dell system.My main computer activity is in the
> areas of family tree, and documentation thereof.Currently, I am converting
> a life time of pictures to digital form as a part of thiseffort.In as much
> as I do know COBOL, and have taught it in the company many years ago,but
> have kept up on the CLC activities, I feel that there are some simple
> thingsI would like to do just to be able to show off a little, and these
> are mainly inwhat I would call utilites. I am not happy with the idea of
> using Windows xpbut I believe I could shoe horn some simple things into the
> system, that wouldmake me happier, and perhaps shut me up<G>

Well, I wrote my first COBOL program in 1968 so you know where I am coming 
from. I use Linux but I have a Windows 2000 partition I seldom use. I will go 
through hte drill of installing one or both COBOLs on that partition and make 
notes as I go. Then I can pass all this lore on to you. Please be patient 
because this is a busy time for me. 

I am reasonably sure that your rig is more than adequate.

Best wishes for the holidays and thanks for writing.
-- 
John Culleton
Able Indexing and Typesetting
Precision typesetting (tm) at reasonable cost.
Satisfaction guaranteed. 
http://wexfordpress.com

[Tiny-cobol-users] Open source COBOL G/L?

[John R. C. <jo...@we...>]

Eons ago (well about 20 years ago) I developed my own version of a G/L system 
in RM COBOL and offered it for sale even. Well, I have lost all the source 
code and don't have the energy or perhaps even the skill level to start all 
over. I have been looking at some freeware/shareware GL packages but they all 
have installation headaches and are written in Perl or php with a Postgres or 
MySql backend etc. etc.

I am looking for the eqivalent in COBOL, for free or close to free. I don't 
have a problem with adapting someone else's code to my particular flavor of 
COBOL or my particular needs. But I prefer not to reinvent the entire wheel 
once again.  

Is there any hope or should I just climb down in the trench and start digging?
I have both O-C and T-C up and running on my Linux Slackware ver 11 system. 
-- 
John Culleton
Able Indexing and Typesetting
Precision typesetting (tm) at reasonable cost.
Satisfaction guaranteed. 
http://wexfordpress.com

Re: [Tiny-cobol-users] Tiny-cobol-users Digest, Vol 6, Issue 3

[Rod S. <rod...@ya...>]

David you are correct as usual.  I did install FC6
with the 64 bit architecture.  Thanks.  I will
reinstall FG6 with 32 bit and try again.


--- tin...@li...
wrote:

> Send Tiny-cobol-users mailing list submissions to
> 	tin...@li...
> 
> To subscribe or unsubscribe via the World Wide Web,
> visit
> 
>
https://lists.sourceforge.net/lists/listinfo/tiny-cobol-users
> or, via email, send a message with subject or body
> 'help' to
> 	tin...@li...
> 
> You can reach the person managing the list at
> 	tin...@li...
> 
> When replying, please edit your Subject line so it
> is more specific
> than "Re: Contents of Tiny-cobol-users digest..."
> 
> 
> Today's Topics:
> 
>    1. Re: Assembler messages: (David Essex)
> 
> 
>
----------------------------------------------------------------------
> 
> Message: 1
> Date: Fri, 17 Nov 2006 20:28:28 -0500
> From: David Essex <de...@us...>
> Subject: Re: [Tiny-cobol-users] Assembler messages:
> To: tiny-cobol-users
> <tin...@li...>
> Message-ID: <455...@us...>
> Content-Type: text/plain; charset=us-ascii;
> format=flowed
> 
> Rod Schell wrote:
> 
>  > I just upgraded to FC6 and I get this message
> when I
>  > compile my cobol programs.
>  >
>  > test3.s:10:
>  > Error: suffix or operands invalid for 'push'
>  > ...
>  >
>  > I am not sure how to resolve.
> 
> I think Fedora Core 6 uses GCC 4.x.
> AFIK, no TC testing has been done with GCC 4.x.
> 
> BTW, are you using a 64-bit machine ?
> TC only generates 32-bit assembler code, and there
> are no plants to port
> TC to any x86 64-bit architecture.
> 
> In any case one would need a small sample of both
> the source COBOL code,
> and the generated GNU assembler code, to check for
> any code generation
> problems.
> 
> htcobol -S test3.cob
> 
> 
> 
> 
> 
> 
> ------------------------------
> 
>
-------------------------------------------------------------------------
> Take Surveys. Earn Cash. Influence the Future of IT
> Join SourceForge.net's Techsay panel and you'll get
> the chance to share your
> opinions on IT & business topics through brief
> surveys - and earn cash
>
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
> 
> ------------------------------
> 
> _______________________________________________
> Tiny-cobol-users mailing list
> Tin...@li...
>
https://lists.sourceforge.net/lists/listinfo/tiny-cobol-users
> 
> 
> End of Tiny-cobol-users Digest, Vol 6, Issue 3
> **********************************************
> 



 
____________________________________________________________________________________
Sponsored Link

Compare mortgage rates for today. 
Get up to 5 free quotes. 
Www2.nextag.com

Re: [Tiny-cobol-users] Assembler messages:

[David E. <de...@us...>]

Rod Schell wrote:

 > I just upgraded to FC6 and I get this message when I
 > compile my cobol programs.
 >
 > test3.s:10:
 > Error: suffix or operands invalid for 'push'
 > ...
 >
 > I am not sure how to resolve.

I think Fedora Core 6 uses GCC 4.x.
AFIK, no TC testing has been done with GCC 4.x.

BTW, are you using a 64-bit machine ?
TC only generates 32-bit assembler code, and there are no plants to port
TC to any x86 64-bit architecture.

In any case one would need a small sample of both the source COBOL code,
and the generated GNU assembler code, to check for any code generation
problems.

htcobol -S test3.cob

Re: [Tiny-cobol-users] Assembler messages:

[Merlin M. <mmo...@gm...>]

On 11/17/06, John R. Culleton <jo...@we...> wrote:
> On Thursday 16 November 2006 22:30, Rod Schell wrote:
> > I just upgraded to FC6 and I get this message when I
> > compile my cobol programs.
> >
> What is FC6? And how does it relate to Tiny-COBOL?

Fedora Core, a recent variant of Redhat linux.   He is having trouble
compiling the above assembly files.

merlin

Re: [Tiny-cobol-users] Assembler messages:

[John R. C. <jo...@we...>]

On Thursday 16 November 2006 22:30, Rod Schell wrote:
> I just upgraded to FC6 and I get this message when I
> compile my cobol programs.
>
What is FC6? And how does it relate to Tiny-COBOL?

You might do better on some more generic COBOL newsgroups.



-- 
John Culleton
Able Indexing and Typesetting
Precision typesetting (tm) at reasonable cost.
Satisfaction guaranteed. 
http://wexfordpress.com

[Tiny-cobol-users] Assembler messages:

[Rod S. <rod...@ya...>]

I just upgraded to FC6 and I get this message when I
compile my cobol programs.

test3.s:10: Error: suffix or operands invalid for
`push'
test3.s:15: Error: suffix or operands invalid for
`push'
test3.s:16: Error: suffix or operands invalid for
`push'
test3.s:18: Error: suffix or operands invalid for
`push'
test3.s:26: Error: suffix or operands invalid for
`push'

I am not sure how to resolve.


 
____________________________________________________________________________________
Sponsored Link

Mortgage rates near 39yr lows. 
$310k for $999/mo. Calculate new payment! 
www.LowerMyBills.com/lre

[Tiny-cobol-users] COBOL calling Java

[GeraldT <ja...@ya...>]

Hi guys,

I'm not sure if this is the right place to ask this question but please do
direct me somwhere else if I am. Anyway, does anyone know how to call a Java
program from COBOL? Specifically microfocus Cobol. Thanks!
-- 
View this message in context: http://www.nabble.com/COBOL-calling-Java-tf2566640.html#a7153757
Sent from the tiny-cobol-users mailing list archive at Nabble.com.

[Tiny-cobol-users] Problem with flex - scan.c compile error

[David E. <de...@us...>]

There is a macro 'define' problem with flex versions greater that 2.5.4.

This will cause TC to fail to compile with the following error message.

  > ...
  > gcc ... -c scan.c
  > scan.c:1062: parse error before `YY_PROTO'
  > ...

This is caused by the following macro 'define', as found in scan.l:61.

#define YY_DECL int yylex2 YY_PROTO((void))

The following code will fix the problem for all version (I hope) of flex.

#ifdef YY_USE_PROTOS
#define YY_DECL int yylex2 YY_PROTO((void))
#else
#define YY_PROTO(proto) (void)
#define YY_DECL int yylex2(void)
#endif


Hope this helps.

David Essex

[Tiny-cobol-users] Building Tiny Cobol

[Hopetoun T. <ho...@co...>]

Hi

A couple of people have had trouble compiling tinycobol on their =
systems.  I was one of them, however after much trial and error I have =
successfully built tinycobol on my Ubuntu system.

You MUST download and install the berkley db library 1.85.4 and the =
yacc-1.9.3-1 packages along with the tinycobol package.

I found that you must remove the version of flex that comes with your =
installation and install flex-old.  I used flex-old 2.5.4a-7.

then in the tinycobol source code directory type

./configure --with-yacc=3Dyacc193

then I generally do=20

make=20

then=20

make install

you will still get some warnings but they shouldn't matter.

regards
Kerry Hotker

[Tiny-cobol-users] bug:not insert negative in variable

[Carlucio L. <car...@te...>]

hi,


       PROGRAM-ID.    bug2.
       AUTHOR. Carlucio Lopes.
       ENVIRONMENT DIVISION.
       DATA DIVISION.
       WORKING-STORAGE SECTION.
       77 SEQ-77  PIC S99V99 VALUE ZEROS.
       77 SEQ-ED  PIC -ZZ,99.
       PROCEDURE DIVISION.
       INICIO.
           DISPLAY "Accept Negative value=" LINE 10 POSITION 10
           ACCEPT SEQ-77 LINE 10 POSITION 35
           MOVE SEQ-77 TO SEQ-ED
           DISPLAY SEQ-ED LINE 10 POSITION 45
           stop " "
           STOP RUN.


TinyCOBOL - pre alpha 0.62.11 (linux-gnu 2005/11/08)
(C) 1991, 1993, 1999-2005 Rildo Pragana et. al.

thank
-- 
Carlucio Lopes-Cobolito 62-8415-8939 3526-8001 msn:car...@ho...
Entenda Pirataria --> http://www.carlinux.v10.com.br
Compilador Cobol Free http://tinycobol.org em Portugues
Porque Cobol? --> http://www.clubecobol.com.br/cc_porque_cobol_02.asp
Projeto Debian Brasil http://cdd.debian-br.org
Tinycobol/Debian Gnu-Linux/Postgresql/Tcl-Tk

Re: [Tiny-cobol-users] Building TC 0.63

[Carlucio L. <car...@te...>]

On Thu, 14 Sep 2006 10:41:54 +0200
"Volker Englisch" <co...@wo...> wrote:

> 
> Making all in compiler
> make[1]: Entering directory `/home/e/eh41/tinycobol-0.63/compiler'
> ~/bin/gcc -I/usr/include -I/usr/local/include -I../lib -I../ -c scan.c
> scan.c:1062: parse error before `YY_PROTO'

install flex-old in Debian:

carlucio@carlinux1:~$ dpkg -l|grep flex
rc  flex           2.5.31-31      A fast lexical analyzer generator.
ii  flex-old       2.5.4a-6       The old version of the fast lexical
analyzer

carlucio@carlinux1:~$ dpkg -l|grep bison
ii  bison-1.35     1.35-4         A parser generator that is compatible
with Y

-- 
Carlucio Lopes-Cobolito 62-8415-8939 3526-8001 msn:car...@ho...
Entenda Pirataria --> http://www.carlinux.v10.com.br
Compilador Cobol Free http://tinycobol.org em Portugues
Porque Cobol? --> http://www.clubecobol.com.br/cc_porque_cobol_02.asp
Projeto Debian Brasil http://cdd.debian-br.org
Tinycobol/Debian Gnu-Linux/Postgresql/Tcl-Tk

[Tiny-cobol-users] Building TC 0.63

[Volker E. <co...@wo...>]

Hi!

I'm successlessly trying to build TC 0.63 on a Debian system. After some time seeking for information what might go wrong, I installed yacc193 from the TC download area, as well as gcc295 as a replacement for gcc335 on this system. Neither lead to success. Again seeking for information, I only found some in Portugues language, what I'm not capable of reading.

So, the error list when doing a "make" starts as follows:

Making all in compiler
make[1]: Entering directory `/home/e/eh41/tinycobol-0.63/compiler'
~/bin/gcc -I/usr/include -I/usr/local/include -I../lib -I../ -c scan.c
scan.c:1062: parse error before `YY_PROTO'
scan.c:1065: register name not specified for `yy_cp'
scan.c:1065: register name not specified for `yy_bp'
scan.c:1066: register name not specified for `yy_act'
scan.l:122: parse error before `switch'
scan.l:125: redefinition of `yy_start'
scan.c:272: `yy_start' previously defined here
scan.l:125: warning: data definition has no type or storage class
scan.l:126: parse error before `break'
scan.l:129: redefinition of `yy_start'
scan.l:125: `yy_start' previously defined here
scan.l:129: warning: data definition has no type or storage class

[... and so on]

The configuration process did not show any error. Maybe someone on this list could enlighten me what I could do in the next step to get TC compiled?

TIA
Volker
-- 
"Feel free" - 10 GB Mailbox, 100 FreeSMS/Monat ...
Jetzt GMX TopMail testen: http://www.gmx.net/de/go/topmail

Re: [Tiny-cobol-users] Any one play wth COBOL as a cgi language?

[David E. <de...@us...>]

John R. Culleton wrote:

  > I work on Linux and have both tiny COBOL and
  > Open COBOL on my system.
  > I know that there have been efforts in the area of
  > interfacing to Tcl/Tk.
  > But what about html?

An example on how to interface TC with CGI, can be found in the
'test.code/tdb03' directory.

In this example TC does not interact with CGI directly, but uses calls
to a small C library 'cgi-util'.

The sample program also interfaces with a postgresql SQL database.

Unfortunatly, I do not have a postgresql installed, and could not test
the sample application. However it should work.

Note that all the elements for a typical CGI application are present in
the sample, so it could be used as a template.

Hope this helps.

Re: [Tiny-cobol-users] Any one play wth COBOL as a cgi language?

[Merlin M. <mmo...@gm...>]

On 9/12/06, John R. Culleton <jo...@we...> wrote:
> I work on Linux and have both tiny COBOL and Open COBOL on my
> system. I know that there have been efforts in the area of interfacing
> to Tcl/Tk. But what about html?

I did some cgi programming with acucobol to hook up our erp to the
PDAs used by people on the floor.  However, once we ported the app to
use postgresql as a backend, we moved all the web stuff (there wasn't
much) to php.

merlin

[Tiny-cobol-users] Any one play wth COBOL as a cgi language?

[John R. C. <jo...@we...>]

I work on Linux and have both tiny COBOL and Open COBOL on my
system. I know that there have been efforts in the area of interfacing
to Tcl/Tk. But what about html?

-- 
John Culleton
Able Indexing and Typesetting
Precision typesetting (tm) at reasonable cost.
Satisfaction guaranteed. 
http://wexfordpress.com

[Tiny-cobol-users] Building TC with GCC 3.3.5

[<co...@wo...>]

Hi!

I'm actually trying to install TC 0.63 on a new Debian box using GCC 3.3.5. It doesn't build, instead it fails with messages like:

scan.c:1060: error: syntax error before "YY_PROTO"
scan.l:120: error: syntax error before "switch"
scan.l:123: error: redefinition of `yy_start"
scan.c:272: error: `yy_start" previously defined here
scan.l:123: Warnung: data definition has no type or storage class

... and so on. In the mailing list archive, I found that someone asked the same question some time ago, but I didn't see any answers.

So, how to build the tarball on this system? Any hint?

TIA
Volker

-- 


"Feel free" – 10 GB Mailbox, 100 FreeSMS/Monat ...
Jetzt GMX TopMail testen: http://www.gmx.net/de/go/topmail

[Tiny-cobol-users] RES: Como instalar o TC no Windows com MinGW

[Jose A. R. <Jos...@ba...>]

Translating:
=20
Hi folks,
=20
I am trying to install Tiny Cobol on MinGw at Windows 2000, but after =
many tryings I do not have done it yet.
=20
I would like to know which packs I need download, the order of =
instalation e where to set that files which does not have rpm or deb .
=20
I am totally newbie in using Linux and GNU tools.
=20
Hugs.
=20
Ed.

	-----Mensagem original-----
	De: tin...@li... =
[mailto:tin...@li...] Em nome de Ednei =
Soares
	Enviada em: quinta-feira, 31 de agosto de 2006 15:39
	Para: tin...@li...
	Assunto: [Tiny-cobol-users] Como instalar o TC no Windows com MinGW
=09
=09
	Ol=E1 Pessoal,

	Estou tentando instalar o TC com MinGw no Windows 2000, mas ap=F3s =
v=E1rias tentativas ainda n=E3o consegui.

	Gostaria de saber quais pacotes eu preciso baixar, qual a ordem =
instala=E7=E3o e onde colocar aqueles arquivos que n=E3o t=EAm =
instalador.

	Sou totalmente leigo no uso do Linux e de ferramentas GNU.

	Abra=E7os,

	Ed

[Tiny-cobol-users] Como instalar o TC no Windows com MinGW

[Ednei S. <soa...@gm...>]

Ol=E1 Pessoal,

Estou tentando instalar o TC com MinGw no Windows 2000, mas ap=F3s v=E1rias
tentativas ainda n=E3o consegui.

Gostaria de saber quais pacotes eu preciso baixar, qual a ordem instala=E7=
=E3o e
onde colocar aqueles arquivos que n=E3o t=EAm instalador.

Sou totalmente leigo no uso do Linux e de ferramentas GNU.

Abra=E7os,

Ed

Re: [Tiny-cobol-users] PROGRAM ID

[David E. <de...@ar...>]

Roger While wrote:
  > David Essex wrote :

  > ...
  >> BTW, doesn't the above methodology lead to illegal program
  >> name lengths under certain conditions ?
  >
  > Haven't come across one yet :-)
  > Remember COBOL words are restricted to 31 characters
  > (also in 2002/2008 standards) so should not be a problem.
  > Even if all 31 characters were mangled, 93 characters for
  > gcc is no problem :-)

Some mainframe COBOL compilers, as an extension, support names as long
as 160 characters.


  >> Anyway, I think the main issue here is program naming conventions
  >> are platform (OS ?) dependent.
  >> ...
  >> Especially when a COBOL 'module' is called from from another
  >> language. So basically you lose some OS functionality and add
  >> some more complexity.
  >
  > Well that's always going to be the case. If you want to call
  > COBOL from non-COBOL, then you are always going to have to look
  > at the particular COBOL implementation as to how to do this.

IMHO, that should not be the case.

On many platforms you can mix C, C++ and/or Pascal (and many other
languages) without too much difficulty, regardless of the compiler. And
certainly does not require a special link-loader just to resolve the name.

Perhaps a special C macro, much like when a C programs calls C++
functions (#ifdef _cplusplus ...).
This could perform some form of 'name (un)mangling', if required.
This should work, I think.


  > ...
  > The 2002 standard has :
  > PROGRAM-ID. program-name [AS literal] ......

What is the point of having two program-names ?

Anyway, water under the bridge, unless some else would like
to create a new TC release.

[Tiny-cobol-users] PROGRAM ID

[Roger W. <si...@si...>]

>
>   > David Essex wrote :
>   >> The "-" is considered an invalid character, for basically
>   >> two reasons.
>   >>
>   >> First, according to the COBOL-85 standard, if memory serves
>   >> me, the program-id (or program name) is implementation defined.
>   >
>   > Wrong.
>   > The program name should conform to the rule for formation
>   > of a user-defined word. The "-" is allowed with the constraint
>   > that it may not appear as the first or last character. (ANS85)
>   > The 2002/2008 standards go further in allowing the underline
>   > "_" (not as first or last character) and "extended letters".
>
>It think you may be correct on this, assuming your interpretation of
>'user-defined word' applies to 'program-name'.
>
>However, in DOS/VSE installations I've worked with, the 'program-name'
>was limited to 8 characters with no hyphens. At least I've never
>encountered a program-name with a "-" character.
>
>Also most modern COBOL compilers allow the use of quoted names, or
>literals, for program names.
>


Not maybe, I AM correct on this.
All the standards define program-name as category user-defined-word.


>   >> Secondly, since TC converts the COBOL program to x86
>   >> (GNU) assembler, it limits what characters can be used.
>   >> Unfortunately the '-' character is interpreted by x86 (GNU)
>   >> assembler as a minus sign. And character substitution,
>   >> such as '$', leads to other problems.
>   >
>   > Yep, same problem in OC with "C".
>   > What we do in OC is replace invalid C/asm characters
>   > with an underline followed by a 2 character hex representation
>   > of the character.
>   >
>   > Also note the same problem with a numeric (0 ... 9) as the first
>   > character which is legal in Cobol.
>
>BTW, doesn't the above methodology lead to illegal program name lengths
>under certain conditions ?
>


Haven't come across one yet :-)
Remember Cobol words are restricted to 31 characters
(also in 2002/2008 standards) so should not be a problem.
Even if all 31 characters were mangled, 93 characters for
gcc is no problem :-)

>Anyway, I think the main issue here is program naming conventions are
>platform (OS ?) dependent.
>
>Yes, one possible solution to the use of 'illegal characters'
>in program names is 'name mangling'. However doing so
>requires some way of automatically resolving the program name.

Which we do in OC; as do other Cobol compilers/runtimes.

>Especially when a COBOL 'module' is called from from another language.
>So basically you lose some OS functionality and add some more complexity.
>

Well that's always going to be the case. If you want to call Cobol from
non-Cobol, then you are always going to have to look at the particular
Cobol implemetation as to how to do this.


>TC, being a x86 platform only compiler, resolved this issue by dropping
>some COBOL standard compliance, in favor of more platform functionality.
>Basically a COBOL program can be viewed as just another C function.
>
>Perhaps a better solution for TC would have been as follows.
>
>PROGRAM-ID. { user-defined-word | literal } ...
>
>If a user-defined-word is used, then some form of 'name mangling' is
>performed (if required).
>If a literal is used, only platform specific legal characters are permitted.
>

The 2002 standard has :
PROGRAM-ID. program-name [AS literal] ......


>Anyway my two cents worth.

Indeed!  :-)

Roger

[Tiny-cobol-users] Error while compiling

[el_bingeroso <fel...@gm...>]

Good evening everyone,
i've recently tryed to compile a hello world program named hello.cob
proceeded to htcobol hello.cob
then 
as -o hello.o hello.s
and receveid the following message
Assembly messages
Error
Can't open hello.s for reading
hello.s : no errors

I have gcc the libraries installed and nasm too. What's going on here?
Thanks for everything.
-- 
View this message in context: http://www.nabble.com/Error-while-compiling-tf2161519.html#a5973707
Sent from the tiny-cobol-users forum at Nabble.com.

Re: [Tiny-cobol-users] PROGRAM ID

[David E. <de...@ar...>]

Roger While wrote:

  > David Essex wrote :
  >> The "-" is considered an invalid character, for basically
  >> two reasons.
  >>
  >> First, according to the COBOL-85 standard, if memory serves
  >> me, the program-id (or program name) is implementation defined.
  >
  > Wrong.
  > The program name should conform to the rule for formation
  > of a user-defined word. The "-" is allowed with the constraint
  > that it may not appear as the first or last character. (ANS85)
  > The 2002/2008 standards go further in allowing the underline
  > "_" (not as first or last character) and "extended letters".

It think you may be correct on this, assuming your interpretation of
'user-defined word' applies to 'program-name'.

However, in DOS/VSE installations I've worked with, the 'program-name'
was limited to 8 characters with no hyphens. At least I've never
encountered a program-name with a "-" character.

Also most modern COBOL compilers allow the use of quoted names, or
literals, for program names.

  >> Secondly, since TC converts the COBOL program to x86
  >> (GNU) assembler, it limits what characters can be used.
  >> Unfortunately the '-' character is interpreted by x86 (GNU)
  >> assembler as a minus sign. And character substitution,
  >> such as '$', leads to other problems.
  >
  > Yep, same problem in OC with "C".
  > What we do in OC is replace invalid C/asm characters
  > with an underline followed by a 2 character hex representation
  > of the character.
  >
  > Also note the same problem with a numeric (0 ... 9) as the first
  > character which is legal in Cobol.

BTW, doesn't the above methodology lead to illegal program name lengths
under certain conditions ?

Anyway, I think the main issue here is program naming conventions are
platform (OS ?) dependent.

Yes, one possible solution to the use of 'illegal characters'
in program names is 'name mangling'. However doing so
requires some way of automatically resolving the program name.
Especially when a COBOL 'module' is called from from another language.
So basically you lose some OS functionality and add some more complexity.

TC, being a x86 platform only compiler, resolved this issue by dropping
some COBOL standard compliance, in favor of more platform functionality.
Basically a COBOL program can be viewed as just another C function.

Perhaps a better solution for TC would have been as follows.

PROGRAM-ID. { user-defined-word | literal } ...

If a user-defined-word is used, then some form of 'name mangling' is
performed (if required).
If a literal is used, only platform specific legal characters are permitted.

Anyway my two cents worth.

[Tiny-cobol-users] PROGRAM ID

[Roger W. <si...@si...>]

David wrote :

>The "-" is considered an invalid character, for basically two reasons.
>
>First, according to the COBOL-85 standard, if memory serves me, the
>program-id (or program name) is implementation defined.

Wrong.
The program name should conform to the rule for formation
of a user-defined word. The "-" is allowed with the constraint
that it may not appear as the first or last character. (ANS85)
The 2002/2008 standards go further in allowing the
underline "_" (not as first or last character) and
"extended letters".


>Secondly, since TC converts the COBOL program to x86 (GNU) assembler, it
>limits what characters can be used.
>Unfortunately the '-' character is interpreted by x86 (GNU) assembler as
>a minus sign. And character substitution, such as '$', leads to other
>problems.

Yep, same problem in OC with "C".
What we do in OC is replace invalid C/asm characters
with an underline followed by a 2 character hex representation
of the character.

Also note the same problem with a numeric (0 ... 9) as the first
character which is legal in Cobol.

Roger While

Re: [Tiny-cobol-users] PROGRAM ID

[David E. <de...@ar...>]

Bill Gunshannon wrote:

  > Not a real problem, but a question out of curiosity.
  >
  > I just tried to compile a program I already had and
  > it flagged all the "-" in the PROGRAM ID as errors.
  > I could not find anywhere in any of my COBOL books
  > where this is not allowed.
  > As a matter of fact, I have a Hardcover book called
  > the ANSI COBOL Programmer Companion (or somethign close
  > to that) that specifically said that "-" is a legal
  > character in the PROGRAM ID.
  >
  > Comments anyone?  Did the standard change?
  > This program compiles just fine with the compiler
  > under OpenVMS.

The "-" is considered an invalid character, for basically two reasons.

First, according to the COBOL-85 standard, if memory serves me, the
program-id (or program name) is implementation defined.

Secondly, since TC converts the COBOL program to x86 (GNU) assembler, it
limits what characters can be used.
Unfortunately the '-' character is interpreted by x86 (GNU) assembler as
a minus sign. And character substitution, such as '$', leads to other
problems.

So avoid any potential problems, the TC program-id conforms to the C
naming convention.