Re: GNAT on Win 95

Re: GNAT on Win 95

[Tom Griest]

somf@crosslink.net (somf) writes:
>1)  Is the assumption that all Win 95 GNAT ada development will occur from
>the MSDOS emulator command line ?  I wish there was a help capability in
>this emulator.

winhelp gcc

should work for you from the command line.

>2)  What is the SH.exe file used for ?

It is a Unix shell that can be used to run unix script files.
It is often used by the make commands when building GNAT.

>3) One of the docs is in a TEX format.  What is this and how can I convert
>to something windows 95 apps can understand ?

It is a format for a common word processor.  You can view the contents
of gnatdoc.tex with an wordpad, but it is quite old and probably just 
of historical value at this point.

>4)  The readme calls for running SETPATH.BAT first,  then suggests cd to
>the examples subdirectory and doing a make.  The make successfully
>compiles the test_cl file but fails in the bind with several errrors.  The
>errors say the many files need to be recompiled.  Mostly, because
>system.ads was modifed, it says.   I tried recompiling the files by hand,
>but I am not sure where the results of these compilations are supposed to
>go.  Is there a global make file ?  Why is this happening ?  Is there
>something else that could be wrong ? 

There is an error in the MS C library that is used in timestamping
the files.  This results in the modify times of files actually changing
when switching to/from daylight savings time.  This has been remedied
in the 3.04 release of GNAT/Win32 by bypassing the C library and going
directly to the Win32 interface which does not exhibit the error.
3.04 is not out yet for Win32, but should be on the net VSN.

There are also a set of updates which fix this problem and a couple
others and these can be found in:
    ftp.cs.yale.edu/pub/gnat/updates/gnat301a

If you get these, it should fix you difficulties.

-Tom

GNAT on Win 95

[somf]

1)  Is the assumption that all Win 95 GNAT ada development will occur from
the MSDOS emulator command line ?  I wish there was a help capability in
this emulator.

2)  What is the SH.exe file used for ?

3) One of the docs is in a TEX format.  What is this and how can I convert
to something windows 95 apps can understand ?

4)  The readme calls for running SETPATH.BAT first,  then suggests cd to
the examples subdirectory and doing a make.  The make successfully
compiles the test_cl file but fails in the bind with several errrors.  The
errors say the many files need to be recompiled.  Mostly, because
system.ads was modifed, it says.   I tried recompiling the files by hand,
but I am not sure where the results of these compilations are supposed to
go.  Is there a global make file ?  Why is this happening ?  Is there
something else that could be wrong ? 

thanks in advance to anyone that can help with these questions.

Re: GNAT on Win 95

[Philip Brashear]

In article <somf-3005960732350001@dyn25b.springfield.va.us.crosslink.net>,
somf <somf@crosslink.net> wrote:
...
>4)  The readme calls for running SETPATH.BAT first,  then suggests cd to
>the examples subdirectory and doing a make.  The make successfully
>compiles the test_cl file but fails in the bind with several errrors.  The
>errors say the many files need to be recompiled.  Mostly, because
>system.ads was modifed, it says.   I tried recompiling the files by hand,
>but I am not sure where the results of these compilations are supposed to
>go.  Is there a global make file ?  Why is this happening ?  Is there
>something else that could be wrong ? 
>
>thanks in advance to anyone that can help with these questions.


Go to ftp.cs.yale.edu, work down the tree to the gnat directory, and look
in "updates".  The one you want is the last one -- it's a fix to the
Daylight Savings Time problem that Microsoft apparently created.  That will
fix you "recompile" problem (though you may have to re-install).

I'm using the Win 95 version of GNAT on a daily basis.  So far, there's
one rather obscure bug (related to the use of a discriminant to control
the length of a string) that's keeping me from running one of my tools.
I know that bug was fixed in the 3.03 version of GNAT that runs under DOS
and UNIX; however, the Win 95 version seems to be stalled at 3.01a.
(Meanwhile, the other ports are up to 3.04, and I'm getting frustrated.)

Good Luck!

Phil Brashear
CTA INCORPORATED

Re: GNAT on Win 95

[Andy Wilson]

somf wrote:
> 
> 1)  Is the assumption that all Win 95 GNAT ada development will occur from
> the MSDOS emulator command line ?  I wish there was a help capability in
> this emulator.
> 
> 2)  What is the SH.exe file used for ?
> 
> 3) One of the docs is in a TEX format.  What is this and how can I convert
> to something windows 95 apps can understand ?
> 
> 4)  The readme calls for running SETPATH.BAT first,  then suggests cd to
> the examples subdirectory and doing a make.  The make successfully
> compiles the test_cl file but fails in the bind with several errrors.  The
> errors say the many files need to be recompiled.  Mostly, because
> system.ads was modifed, it says.   I tried recompiling the files by hand,
> but I am not sure where the results of these compilations are supposed to
> go.  Is there a global make file ?  Why is this happening ?  Is there
> something else that could be wrong ?
> 
> thanks in advance to anyone that can help with these questions.

Hi.
I tried GNAT301a on Win95 recently too. Whilst I've not used it in earnest 
yet, I found the same problem with the "system.ads" needing recompiling.
To get over this I tried re-"make"-ing the library. 
cd to the "adainclude" directory. There you'll find a makefile. Have a look 
in this and it tells you about recompiling the library and what to do with 
the compiled *.obj etc.. Then just type "make" in this directory. When 
finished,  copy the *.obj and *.ali(? i think - check iwhat it says in the 
makefile) to the lib directory. This worked fine for me.

Andy Wilson
wilsona@logica.com
http://public.logica.com/~wilsona

Re: GNAT on Win 95

[Bob Crispen]

Andy Wilson (wilsona@logica.com) sez:

>I tried GNAT301a on Win95 recently too. Whilst I've not used it in earnest
>yet, I found the same problem with the "system.ads" needing recompiling.
>To get over this I tried re-"make"-ing the library.
>cd to the "adainclude" directory. There you'll find a makefile. Have a look
>in this and it tells you about recompiling the library and what to do with
>the compiled *.obj etc.. Then just type "make" in this directory. When
>finished,  copy the *.obj and *.ali(? i think - check iwhat it says in the
>makefile) to the lib directory. This worked fine for me.

I don't recall how I came to know this (perhaps it was from reading
the documents, but that's highly unlikely ;-) but there's a mailing
list where Tom Griest sends announcements about Gnat for Win32.
For example, last night I got the announcement of the availability
of 3.04a.

What you experienced was a bug that there was already a patch out there
for.  I'm glad the recompile went fine, but it really wasn't necessary.

Bob Crispen
revbob@eight-ball.hv.boeing.com
Speaking for myself, not my company

Re: GNAT on Win 95

[Theodore E. Dennison]

Bob Crispen wrote:
> 
> I don't recall how I came to know this (perhaps it was from reading
> the documents, but that's highly unlikely ;-) but there's a mailing
> list where Tom Griest sends announcements about Gnat for Win32.
> For example, last night I got the announcement of the availability
> of 3.04a.

Availability where? Its not at cs.nyu.edu (at least not in any obvious
place).

> What you experienced was a bug that there was already a patch out there
> for.  I'm glad the recompile went fine, but it really wasn't necessary.

Patch? That's not at cs.nyu.edu either. Isn't nyu the "official" source
for gnat distributions? Does Labtek have their own distribution site?

-- 
T.E.D.          
                |  Work - mailto:dennison@escmail.orl.mmc.com  |
                |  Home - mailto:dennison@iag.net              |
                |  URL  - http://www.iag.net/~dennison         |

Re: GNAT on Win 95

[Theodore E. Dennison]

Theodore E. Dennison wrote:
> 
> Bob Crispen wrote:
> >
> > What you experienced was a bug that there was already a patch out there
> > for.  I'm glad the recompile went fine, but it really wasn't necessary.
> 
> Patch? That's not at cs.nyu.edu either. Isn't nyu the "official" source
> for gnat distributions? Does Labtek have their own distribution site?

Ted, you ignorant baboon, its at ftp://ftp.cs.yale.edu/pub/gnat/ !

(Everyone else seems to have so much fun abusing me, I thought I'd
try it myself).

Lycos was kind enough to find it for me. It would be nice if some
of the places that carry or mention Gnat would mention this site 
too.

Interestingly enough, what Lycos actually found for me was a link
to a Bob Crispen page thanking Tom Griest for his efforts with this
port. That's awfully considerate of you, Bob (although the background
with the hearts is most sickening). If I was Tom I'd be quite touched,
...and quite happy that I live over a thousand miles away. :-)

-- 
T.E.D.          
                |  Work - mailto:dennison@escmail.orl.mmc.com  |
                |  Home - mailto:dennison@iag.net              |
                |  URL  - http://www.iag.net/~dennison         |

Re: GNAT on Win 95

[progers]

In <31B87984.2F1CF0FB@escmail.orl.mmc.com>, "Theodore E. Dennison" <dennison@escmail.orl.mmc.com> writes:
>Theodore E. Dennison wrote:

[snip]

>Interestingly enough, what Lycos actually found for me was a link
>to a Bob Crispen page thanking Tom Griest for his efforts with this
>port. That's awfully considerate of you, Bob (although the background
>with the hearts is most sickening). If I was Tom I'd be quite touched,
>.....and quite happy that I live over a thousand miles away. :-)

Nah -- Tom's wife Laura has Bob beat easily.  No contest (and I
haven't even seen Bob:)

pat
---------------
Patrick Rogers
progers@acm.org

Re: GNAT on Win 95

[Kenneth Mays]

Greetings,

Gnat V3.04a for MSDOS is available on ftp://www.gwu.edu/pub/ez2load

It will benefit you to download the entire directory (the
directory contains good source code and an Ada editor). This version 
includes tasking. You will get four main preprocessor files 
(gnat341.zip - gnat344.zip). Don't forget the DJGPP compiler and 
other files!!!!

From my notes, it became available around June 3, 1996. As for the 
Labtek version, I have only seen Gnat V3.01a which uses Win95/NT (I 
used it for awhile before I got V3.04 of GNAT).

Have fun,

Ken Mays

Thought of the day:
One does not gain much by mere cleverness.

Re: GNAT on Win 95

[Robert Dewar]

T.E.D says

"Patch? That's not at cs.nyu.edu either. Isn't nyu the "official" source
for gnat distributions? Does Labtek have their own distribution site?"

In the past Labtek has developed patches on their own which have been
at their site, and not always properly coordinated with the release
at NYU. We are attempting to correct this in the future and avoid
further patches (the 4.01a in the NT release comes from Labtek, who
has in the past used the letter to indicate patch levels, but in future
we hope to be able to coordinate with Labtek to avoid this kind of 
confusion).

Robert Dewar
ACT

Re: GNAT on Win 95

[Robert Dewar]

Kenneth said

"Gnat V3.04a for MSDOS is available on ftp://www.gwu.edu/pub/ez2load"

This is misinformation, the version number for MSDOS is 3.04, not
3.04a (the latter is a designatiuon used by Labtek for the NT version).
The DOS 3.04 distribution is also available at the NYU and other mirror
sites, together with the Ez2load utilities (the latter may not be quite
there yet, but if not, will be there by the time you read this)

Re: GNAT on Win 95

[Tom Griest]

Bob Crispen <revbob@EIGHT-BALL.HV.BOEING.COM> writes:
>progers@ACM.ORG sez:
>
?>>>Interestingly enough, what Lycos actually found for me was a link
?>>>to a Bob Crispen page thanking Tom Griest for his efforts with this
?>>>port. That's awfully considerate of you, Bob (although the background
?>>>with the hearts is most sickening). If I was Tom I'd be quite touched,
?>>>.....and quite happy that I live over a thousand miles away. :-)

PR>>Nah -- Tom's wife Laura has Bob beat easily.  No contest (and I
PR>>haven't even seen Bob:)

BC>You will be pleased to know that, yucky hearts aside, I am extremely
BC>cute.
BC>-----
BC>Getting serious for a moment, I think it's absolutely wonderful news
BC>for our community that Robert Dewar is saying they'll be coordinating
BC>a little better with LabTek.  In the past, it depended on what day you
BC>went to the Gnat site whether or not you'd even be let in on the
BC>existence of the LabTek Windows 95 port.  I appreciate (or at any
BC>rate, believe) that there are some financial interests involved.  This
BC>is an unselfish gesture, and I think it ought to be recognized and
BC>applauded.

We have been coordinating to some degree for quite a while now, 
with some success but it has also been less than perfect.
We are trying to address the issues to provide the best 
solution for GNAT users and to apply our talents in the most 
effective way.

BC>-----
BC>While I'm sort of on the same topic, I hope the relevant folks are
BC>paying attention to the Win32 API that's in the process of coming
BC>out, I believe through the Cygnus GnuWin32 folks.  There's an opportunity
BC>for some synergy here, I believe.  I apologize for not knowing the
BC>name of the author; this represents a tremendous amount of work.

I believe most of their GnuWin32 stuff was developed by Steve Chamberlain,
who I understand is recently no longer with Cygnus Support.  However, 
I guess someone by the name of Geoffrey Noer has taken over.
We have been communicating with them on a fairly regular basis,
(like last week) trying to share information whenever possible.  
Ultimately, we want to migrate to the Cygnus tools for Win32, but 
because they have not supported MS library formats, we have had to 
maintain our own port (which had been done prior to the Cygnus port).
We really want to be interoperable with all the MS tools.

There are several related efforts active, including the Win32Ada
bindings, GNAT, Cygnus, ASIS, IDE work, high-level win32 bindings,
C++ interfacing, Distributed Annex stuff, and a bunch of others.
You could spend all your time "coordinating" if you wanted to.
Still, with each release, the GNAT product line becomes more
sophisticated and more polished.  My hope is that we will
merge efforts whenever it makes sense to do so, to avoid
duplication of work and compatibility problems.

-Tom

Re: GNAT on Win 95

[Bob Crispen]

progers@ACM.ORG sez:

>>Interestingly enough, what Lycos actually found for me was a link
>>to a Bob Crispen page thanking Tom Griest for his efforts with this
>>port. That's awfully considerate of you, Bob (although the background
>>with the hearts is most sickening). If I was Tom I'd be quite touched,
>>.....and quite happy that I live over a thousand miles away. :-)
>
>Nah -- Tom's wife Laura has Bob beat easily.  No contest (and I
>haven't even seen Bob:)

You will be pleased to know that, yucky hearts aside, I am extremely
cute.
-----
Getting serious for a moment, I think it's absolutely wonderful news
for our community that Robert Dewar is saying they'll be coordinating
a little better with LabTek.  In the past, it depended on what day you
went to the Gnat site whether or not you'd even be let in on the
existence of the LabTek Windows 95 port.  I appreciate (or at any
rate, believe) that there are some financial interests involved.  This
is an unselfish gesture, and I think it ought to be recognized and
applauded.
-----
While I'm sort of on the same topic, I hope the relevant folks are
paying attention to the Win32 API that's in the process of coming
out, I believe through the Cygnus GnuWin32 folks.  There's an opportunity
for some synergy here, I believe.  I apologize for not knowing the
name of the author; this represents a tremendous amount of work.
-----
This past weekend I got my VRML page in shape and drastically improved
the links.  There are now screen grabs of all the VRML sites on my
list.  It's at http://hiwaay.net/~crispen/vrml/

Why would I bring this up in an Ada newsgroup?  Because there is quite a
fight going on right now in the VRML (Virtual Reality Modeling Language)
community about the external API.  I have been carrying the spear for
having at least one of the APIs be in Java bytecode, with the not very
well hidden agenda of being able to reuse Ada via AppletMagic.

If there are folks here who have an interest in VRML, and who believe
it might have some value when coupled with Ada, you might want
to join this discussion by sending email to majordomo@wired.com
consisting of the line:

        subscribe www-vrml

As with any net community, they appreciate folks being up to speed a
little bit on the issues.  You might want to check my VRML webpage and
follow down some of the links.  And please do lurk a while, read the
specs and so on before you jump in.

I should point out that this is a technical mailing list.  Problems
with installing a VRML browser on your system and working the controls
and questions about where to get a VRML browser are best asked elsewhere.

Bob Crispen
revbob@eight-ball.hv.boeing.com
Speaking for myself, not my company

Re: GNAT on Win 95

[Theodore E. Dennison]

Bob Crispen wrote:
> While I'm sort of on the same topic, I hope the relevant folks are
> paying attention to the Win32 API that's in the process of coming
> out, I believe through the Cygnus GnuWin32 folks.  There's an opportunity
> for some synergy here, I believe.  I apologize for not knowing the
> name of the author; this represents a tremendous amount of work.

Hmm. I know there are already Win32 bindings for GNAT. I take it you
are referring to some sort of effort to make gcc-native Win32 DLL's 
and .h's? That would indeed be "a tremendous amount of work". This
would make it unnessecary to purchase a Microsoft or Borland 
compiler to do serious development on Windows NT and 95. That would
be a major breakthrough.

I'm unclear about what kind of "synergy" is needed, though. The 
Win32 bindings should work with either, right?


-- 
T.E.D.          
                |  Work - mailto:dennison@escmail.orl.mmc.com  |
                |  Home - mailto:dennison@iag.net              |
                |  URL  - http://www.iag.net/~dennison         |

Re: GNAT on Win 95

[Bob Crispen]

Dale Pontius <pontius@TWONKY.BTV.IBM.COM> asks:

>>While I'm sort of on the same topic, I hope the relevant folks are
>>paying attention to the Win32 API that's in the process of coming
>>out, I believe through the Cygnus GnuWin32 folks.  There's an opportunity
>>for some synergy here, I believe.  I apologize for not knowing the
>>name of the author; this represents a tremendous amount of work.

>How does GNUWin32 compare/relate to WINE?
>
>This is the first I've heard of it. So far I'd heard of WINE and
>Willow, and various Public Windows proposals.

Cygnus GnuWin32 is a Gnu C and C++ compiler, being ported to Windows
95 and Windows NT by Cygnus Software.  Alongside this compiler (and
accompanying Gnu tools) there is an effort, which last I heard was
also being supplied by but not originated by Cygnus, to reconstruct
the Win32 API header files and libraries without reference to any
Microsoft code.  I'll see if I can find the exact source and creator.

In the meantime, are WINE and Willow Win32 API clones, or are they
like V, platform-independent APIs?

Bob Crispen
revbob@eight-ball.hv.boeing.com
Speaking for myself, not my company

Re: GNAT on Win 95

[Tom Griest]

pontius@twonky.btv.ibm.com (Dale Pontius) writes:

someone wrote:
>>While I'm sort of on the same topic, I hope the relevant folks are
>>paying attention to the Win32 API that's in the process of coming
>>out, I believe through the Cygnus GnuWin32 folks.  There's an opportunity
>>for some synergy here, I believe.  I apologize for not knowing the
>>name of the author; this represents a tremendous amount of work.
>>-----
DP>How does GNUWin32 compare/relate to WINE?
DP>
DP>This is the first I've heard of it. So far I'd heard of WINE and
DP>Willow, and various Public Windows proposals.

WINE is a way to _run_ Windows programs, whereas GnuWin32 is a way
to _develop_ Win32 programs.  That is WINE is a way to substitute
Linux for Windows, and GnuWin32 is a way to substitue for a Win32 SDK.
I'm not familiar with Willow, but I've heard of attempts to develop 
something like WINE32.  I really think getting a public WINE32 
implementation would be a good thing.  Although it is hard to tell 
what impact such things will have in the long run, it would provide 
more alternatives in the marketplace.

Although the Win32 API is not as "clean" of a specification as I
would like, the underlying functionality of Win32 is actually 
very good (IMHO).  Given the number of platforms (Intel, MIPS, PowerPC,
and Alpha) and shear number of systems that run Win32 (WinNT & Win95), it 
is a pretty sure bet that it will be a primary interface to program to
for some time to come.

-Tom

Re: GNAT on Win 95

[Dale Pontius]

In article <9606101448.AA26858@eight-ball>,
        Bob Crispen <revbob@EIGHT-BALL.HV.BOEING.COM> writes:
>progers@ACM.ORG sez:
>-----
>While I'm sort of on the same topic, I hope the relevant folks are
>paying attention to the Win32 API that's in the process of coming
>out, I believe through the Cygnus GnuWin32 folks.  There's an opportunity
>for some synergy here, I believe.  I apologize for not knowing the
>name of the author; this represents a tremendous amount of work.
>-----
How does GNUWin32 compare/relate to WINE?

This is the first I've heard of it. So far I'd heard of WINE and
Willow, and various Public Windows proposals.

Dale Pontius
(NOT speaking for IBM)

Re: GNAT on Win 95

[Bob Crispen]

"Theodore E. Dennison" <dennison@ESCMAIL.ORL.MMC.COM> sez:

>> While I'm sort of on the same topic, I hope the relevant folks are
>> paying attention to the Win32 API that's in the process of coming
>> out, I believe through the Cygnus GnuWin32 folks.  There's an opportunity
>> for some synergy here, I believe.  I apologize for not knowing the
>> name of the author; this represents a tremendous amount of work.
>
>Hmm. I know there are already Win32 bindings for GNAT. I take it you
>are referring to some sort of effort to make gcc-native Win32 DLL's
>and .h's? That would indeed be "a tremendous amount of work". This
>would make it unnessecary to purchase a Microsoft or Borland
>compiler to do serious development on Windows NT and 95. That would
>be a major breakthrough.
>
>I'm unclear about what kind of "synergy" is needed, though. The
>Win32 bindings should work with either, right?

The synergy comes from the fact that two people (or perhaps groups
of people) are working on something that has common elements.  If
they were to get together and say "I'll do this; you do that," we
groveling (yet demanding) users might get both of them on our desks
a day or two faster.

Ditto for Cygnus and LabTek -- here we've got two groups of people
(saintly people) writing a gcc for Windows 95 and NT.  Both of them
compile C.  One of them compiles C++.  One of them compiles Ada 95.
It's a virtual certainty that the same problems are being solved twice.
It's also a virtual certainty that every day the two common cores of
the compilers move another millimeter farther apart, and the chances
of forever having to have two compilers on my machine increase another
hundredth of a percent.

Not only that, but (being an ignorant, yet carping outsider) I believe
it is only due to pure, blind luck that I can run my LabTek gnat
compiler under the Gnu tools from Cygnus (e.g., bash).  Will it be
true on the next release (and, btw, great current release, Tom!)?

It's not like Ada was Netscape, where there are billions to be made
and trade secrets make sense.  Even the market for programmers on
Win32 machines isn't that big -- at least Win32 programmers who'll
program in Ada despite the absence of Microsoft Visual Ada.  But
there's enough there for people to make a damn decent living, and
cooperation might help get things to market a little faster, improving
everyone's chances of retiring happily.

Bob Crispen
revbob@eight-ball.hv.boeing.com
Speaking for myself, not my company

Re: GNAT on Win 95

[Dale Pontius]

In article <9606121842.AA16493@eight-ball>,
        Bob Crispen <revbob@EIGHT-BALL.HV.BOEING.COM> writes:
>
>In the meantime, are WINE and Willow Win32 API clones, or are they
>like V, platform-independent APIs?
>
I'm not sure, but I think Willow is a Win32 clone.

WINE is an API-clone to run Windows binaries on Linux. They began
with Win16, but have been adding a lot of Win32 function, lately.

Dale Pontius
(NOT speaking for IBM)

Re: GNAT on Win 95

[Robert Dewar]

iBob Crispen says

"Ditto for Cygnus and LabTek -- here we've got two groups of people
(saintly people) writing a gcc for Windows 95 and NT.  Both of them
compile C.  One of them compiles C++.  One of them compiles Ada 95.
It's a virtual certainty that the same problems are being solved twice.
It's also a virtual certainty that every day the two common cores of
the compilers move another millimeter farther apart, and the chances
of forever having to have two compilers on my machine increase another"

What on earth are you talking about. LabTek is not writing a "gcc for 
Windows 95"! The port of GNAT for NT uses the current release of gcc
from FSF with a few extra patches which represent changes present
in the current development version of FSF gcc which are necessary
for GNAT (the maintanence of the FSF version of gcc, which is what
Labtek -- and all other versions of GNAT -- uses, is under the
control and supervision of Richard Kenner of ACT and NYU)

The Cygnus development of gcc is kept syncrhonized with the FSF
development, and Cygnus fixes and improvements are incorporated
into the FSF release.

Bob, you really ought to take the effort to find more about what
is going on and how gcc works before posting a message which is
so confused. 

"One of them compiles C++"

more massive confusion. gcc is a driver program that can call cc1, gnat1,
or cc1plus etc, which are the actual compilers for the different languages.
Each of these compilers incorporates the backend code generator etc that
is a fundamental part of gcc.

If you get the latest version of the gcc driver from either FSF or Cygnus
then they should be, barring minor revision shifts, identical, and either
of them will be happy to load the C++ compiler or the GNAT compiler.

No problems are being solved twice

Nothing is moving apart

If you want to have C++ and Ada on your machine, you definitely need
two compilers, a C++ compiler and an Ada cmpiler. These can both be
loaded using the gcc driver. You definitely do NOT need two copies of
the gcc driver, so once you set things up properly on your machine,
you have one copy of gcc, which compiles C, C++, or Ada (or for that
matter Chill, or Fortran 77, or Objective C, or ...) depending on the
extension of the file (or the -x switch).

Re: GNAT on Win 95

[Jim Kingdon]

> we groveling (yet demanding) users might get both of them on our desks
> a day or two faster.

Well, have you showed up on Cygnus's or Labtek's door carrying a big
pot of money (or even a small pot of money) and a description of what
you want?

This is going to sound like carping to some, but free software
development is driven by two things--people who volunteer their own
effort to make it happen, and people who pay other people to do so.
As an example, users have been asking for a CVS Win95 port ever since
the NT port came out (August 1995), but all that asking hasn't done
any good.  What has done some good is that someone actually did
something of a port and made it available by web/ftp.  Another thing
which would do some good, but which hasn't happened yet, is if someone
would make/fund an effort to get a Win95 port merged into the main CVS
distribution.

> Ditto for Cygnus and LabTek -- here we've got two groups of people
> (saintly people)

Cygnus might be willing to spend small amounts of effort to secure
sainthood.  But if you want large amounts of effort from them, they
will be looking for more than sainthood (I'll give you a hint--what
they seek is, in the US, a pale shade of green).

Re: GNAT on Win 95

[Pascal Obry]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 1826 bytes --]


dewar@cs.nyu.edu (Robert Dewar) wrote:

>If you want to have C++ and Ada on your machine, you definitely need
>two compilers, a C++ compiler and an Ada cmpiler. These can both be
>loaded using the gcc driver. You definitely do NOT need two copies of
>the gcc driver, so once you set things up properly on your machine,
>you have one copy of gcc, which compiles C, C++, or Ada (or for that
>matter Chill, or Fortran 77, or Objective C, or ...) depending on the
>extension of the file (or the -x switch).

Well I think that Bob Crispen is right here. The problem is that
librairies from Cygnus GCC and Labtek GNAT-Win32 are not compatible.
So even if you can use the same driver GCC you need to have 2 copies
of the librairies, 2 copies of the linker. The reusable part the GCC
driver is only 44k :-)

I have tried to put both compiler into the same directory but I wasn't
able to make them working right... Tom Griest told me that they will
merge sometime with Labtek when librairies becomes compatibles (i.e.
using the Microsoft format).

Pascal.

--|------------------------------------------------------------
--| Pascal Obry                               Team-Ada Member |
--|                                                           |
--| EDF-DER-IPN-SID- Ing�nierie des Syst�mes d'Informations   |
--|                                                           |
--| Bureau G1-010           e-mail: pascal.obry@der.edfgdf.fr |
--| 1 Av G�n�ral de Gaulle  voice : +33-1-47.65.50.91         |
--| 92141 Clamart CEDEX     fax   : +33-1-47.65.50.07         |
--| FRANCE                                                    |
--|------------------------------------------------------------
--|
--|   http://ourworld.compuserve.com/homepages/pascal_obry
--|
--|   "The best way to travel is by means of imagination"

Re: GNAT on Win 95

[Bob Crispen]

Robert Dewar <dewar@CS.NYU.EDU> sez:

>"Ditto for Cygnus and LabTek -- here we've got two groups of people
>(saintly people) writing a gcc for Windows 95 and NT.  Both of them
>compile C.  One of them compiles C++.  One of them compiles Ada 95.
>It's a virtual certainty that the same problems are being solved twice.
>It's also a virtual certainty that every day the two common cores of
>the compilers move another millimeter farther apart, and the chances
>of forever having to have two compilers on my machine increase another"
>
>What on earth are you talking about. LabTek is not writing a "gcc for
>Windows 95"! The port of GNAT for NT uses the current release of gcc
>from FSF with a few extra patches which represent changes present
>in the current development version of FSF gcc which are necessary
>for GNAT>

Oh, very well.  Cygnus and Labtek are separately developing
compilers in the gcc family.  Is that precise enough?  You clearly
knew from your other remarks that I was not alleging that these were
clean-room efforts.

>(the maintanence of the FSF version of gcc, which is what
>Labtek -- and all other versions of GNAT -- uses, is under the
>control and supervision of Richard Kenner of ACT and NYU)
>
>The Cygnus development of gcc is kept syncrhonized with the FSF
>development, and Cygnus fixes and improvements are incorporated
>into the FSF release.

Excellent news.  Perhaps I'm not the only one to whom it is news.
Would you be kind enough to mention where, apart from here and now,
you've actually said that with respect to those two compilers?

>"One of them compiles C++"
>
>more massive confusion. gcc is a driver program that can call cc1, gnat1,
>or cc1plus etc, which are the actual compilers for the different languages.
>Each of these compilers incorporates the backend code generator etc that
>is a fundamental part of gcc.

The release notes for the last version of gnat that I installed on
my Sun mentioned what was required to use the same front end for C++,
which is why I knew it was possible on some platforms.  But since
the gnat release notes for Win95 not only did not mention that but
said that C++ was not supported, my statement becomes a *little*
easier to justify.

>If you get the latest version of the gcc driver from either FSF or Cygnus
>then they should be, barring minor revision shifts, identical, and either
>of them will be happy to load the C++ compiler or the GNAT compiler.
[moved]
>Bob, you really ought to take the effort to find more about what
>is going on and how gcc works before posting a message which is
>so confused.

Well, that "barring minor revision shifts" tells me that you, who
of all people in the world should know, don't know either whether the
current releases of the Cygnus and LabTek compilers will permit that!
That's really too much: scolding a compiler user for not knowing what
you, a compiler developer, do not know.

Bob Crispen
revbob@eight-ball.hv.boeing.com
Speaking for myself, not my company