According to Karl Heuer:
> I need to build the latest gcc and gpc for a Deluxe Distribution.
> The current version of gcc is 2.7.2.3, but gpc-2.0 is expecting
> 2.7.2.1. I think 2.7.2.3 is nearly identical to 2.7.2.1, so maybe
> I can just change gpc's Makefile.in so that it expects 2.7.2.3.
> Do you know if this should work?
This works.
However the current beta version gpc-971001 is already much more
stable than gpc-2.0; I hope that we will be able to release gpc-2.1
soon. You can get it from
ftp://agnes.dida.physik.uni-essen.de/gnu-pascal/beta/gpc-971001.tar.gz
Greetings,
Peter
Dipl.-Phys. Peter Gerwinski, Essen, Germany, free physicist and programmer
peter.gerwinski(a)uni-essen.de - http://home.pages.de/~peter.gerwinski/ [971005]
maintainer GNU Pascal [971001] - http://home.pages.de/~gnu-pascal/ [971005]
(Posted to comp.lang.pascal.ansi-iso and to
the GNU Pascal mailing list, gpc at hut dot fi.)
Hello, Extended Pascal experts!
We need your help to increase the compliance of the
GNU Pascal compiler (GPC) to the ISO-10206 Extended Pascal
standard.
1) We need a test suite to check in how far GNU Pascal does
comply to the EP standard. (Our aim is 100%.) Does anyone
know where to get such a thing?
2) The standard says that not only files, but also other
Pascal entities may be bound to external entities, but
the effect of such a binding is "implementation dependent".
How do other implementations handle this?
(* In case you don't know: GNU Pascal is a portable 32-bit
Pascal compiler. It is Free Software according to the GNU
General Public License. The current beta version (gpc-971001)
supports most of ISO-10206 Extended Pascal as well as most of
Borland Pascal 7.0. For more information, see our WWW pages,
http://home.pages.de/~gnu-pascal/. *)
Thanks in advance,
Peter
Dipl.-Phys. Peter Gerwinski, Essen, Germany, free physicist and programmer
peter.gerwinski(a)uni-essen.de - http://home.pages.de/~peter.gerwinski/ [971005]
maintainer GNU Pascal [971001] - http://home.pages.de/~gnu-pascal/ [971005]
Hiya all,
I have just discovered that the "time bug" which I reported a few
days ago (in DJGPP) is not a bug in the GPC compiler at all. It
it seems to have been caused by a bad parameter passed at the
command line ("-extended-pascal", instead of
"--extended-pascal"), which causes ld.exe to give a warning, and
then produces a buggy .EXE file.
Sorry for leading Peter and Frank on a wild goose chase!
Best regards, The Chief
--------
Dr. Abimbola A. Olowofoyeku (The African Chief)
Email: laa12(a)keele.ac.uk
Homepage: http://ourworld.compuserve.com/homepages/African_Chief/
Author of: Chief's Installer Pro 4.01 for Win16 and Win32:
http://www.simtel.net/pub/simtelnet/win3/install/chief401.zip
The African Chief wrote:
> I was not talking about the date/time format at all. I was
> talking about the actual values that are displayed. I just ran
> the program again, and this is the output (DJGPP, running under
> Windows NT v4.0):
>
> "327680.786432.130875392 851968:2490368:2424832"
>
> That is what I mean by "unintelligible". I don't know how to
> translate these values into something that means something to
> human beings.
Ah, OK, that's something different -- namely a bug... :-(
Of course, there should be no need for any translation, as the
values in the time stamp should be the actual time -- and that's
what they are on my Linux box.
Perhaps it's the same bug that was reported about test\extended\time.pas
in <34396148.5C71(a)ujf-grenoble.fr>, but I'll record this, too. Thanks
for the bug report... :-|
I'll take a look into the RTS and see if I can find something wrong.
However, since I don't see the bug I'm not sure if I can find the
problem...
--
Frank Heckenbach, Erlangen, Germany
fjf(a)gmx.de
http://home.pages.de/~fjf/links.htm
According to Maurice Lombardi:
> > Perhaps it's the same bug that was reported about test\extended\time.pas
> > in <34396148.5C71(a)ujf-grenoble.fr>, but I'll record this, too. Thanks
> > for the bug report... :-|
> No it is different. I have checked that this old bug is still there,
> while the example of the Chief works fine on my machine.
> (All is made in plain DOS/ DJGPP v2.1,
> the old bug did not appear under bash shell in the same machine)
I suspect that this bug is not in GPC but in the DJGPP port of
the UNIX function `gettimeofday'. Maybe the `bash' shell catches
and "repairs" the DOS interrupt? I don't know ...
Maybe we should re-write `_p_gettimestamp' for the MSDOS target
using the interrupt directly?
The other bug reported by The African Chief *might* have the same
reason.
A lot of guessing, sorry ...
Peter
Dipl.-Phys. Peter Gerwinski, Essen, Germany, free physicist and programmer
peter.gerwinski(a)uni-essen.de - http://home.pages.de/~peter.gerwinski/ [971005]
maintainer GNU Pascal [971001] - http://home.pages.de/~gnu-pascal/ [971005]
Hi Javier,
I have successfully installed the Solaris binary on the computers at the
math institute of our university -- exactly on the 20th anniversary of the
computer station. :-)
When the next beta version of GPC comes out, can you please provide me a
binary of this, too? Then I could run the test suite on the sparcs and make
sure there are no new bugs, before we can finally release GPC 2.1.
Regards,
Frank
--
Frank Heckenbach, Erlangen, Germany
fjf(a)gmx.de
http://home.pages.de/~fjf/links.htm
Hi all,
This month's edition of the Developers Review (sister magazine of
the Delphi Magazine) carries an article on GPC, written by (yours
truly!! :-)). I
I have already had an email about it!
Best regards, The Chief
--------
Dr. Abimbola A. Olowofoyeku (The African Chief)
Email: laa12(a)keele.ac.uk
Homepage: http://ourworld.compuserve.com/homepages/African_Chief/
Author of: Chief's Installer Pro 4.01 for Win16 and Win32:
http://www.simtel.net/pub/simtelnet/win3/install/chief401.zip
According to Pismo:
> I have posted in the Borland Delphi group(s)that I use GPC and the
> Delphi command line compiler to write CGIs on Linux and NT that
> ported with very few mods.
:-) Good news!
> Have you considered a GPC newsgroup, and if so do you need somewhere
> to host it? I'll be putting up a NNTP server to spool local
> newsgroups for the ISP where my server is hosted and would be able to
> add a group for GPC.
Can one simply create a new Newsgroup "just like that"? I think that
you can do so only for "alt.*" Newsgroups, and for a GNU Pascal Newsgroup
the "natural" name would be "comp.lang.pascal.gnu". The host would not
be a problem, but doesn't the creation of a new Newsgroup require some
voting etc.? It also would be nice to do it like the DJGPP guys and to
route the mailing list into the Newsgroup and vice versa. If you can
help to solve these problems, it would be highly welcome.
> Also if you want to quote anything I've said in email on the GPC-list
> I had assumed it would probably be forwarded there anyway. I only
> wrote you directly because my posts had been rejected with that error
> I sent you recently.
The error message can be ignored; your messages went through to the
list members. (I am CCing this e-mail to the list and will probably
get the same error message again.)
> Thanks a bunch, and I really appreciate your efforts producing GPC.
You are welcome. :-)
Greetings,
Peter
Dipl.-Phys. Peter Gerwinski, Essen, Germany, free physicist and programmer
peter.gerwinski(a)uni-essen.de - http://home.pages.de/~peter.gerwinski/ [971005]
maintainer GNU Pascal [971001] - http://home.pages.de/~gnu-pascal/ [971005]
The African Chief wrote:
> > program x;
> > var time:TimeStamp;
> > begin
> > GetTimeStamp(time);
> > writeln(time.day,'.',time.month,'.',time.year,' ',
> > time.hour,':',time.minute,':',time.second)
> > end.
>
> The output from this is unintelligible. Is there a way to convert
> the output into something useful?
We won't argue about different date/time formats, will we?
I just used the format that's usual around here. Since the writeln
statement should be self-explanatory, you should have no problems
converting it into a format that you prefer.
Regards,
Frank
--
Frank Heckenbach, Erlangen, Germany
fjf(a)gmx.de
http://home.pages.de/~fjf/links.htm