Ada Compiler for PC?

Ada Compiler for PC?

[emerrif]

I urgently require a good, versatile ADA compiler for my PC,
does anyone know where I can buy or download such a thing?

My thanks to all.

Re: Ada Compiler for PC?

[tmoran]

  What do you mean by 'PC': 640K DOS?, big memory via DOS DPMI?,
Windows 9X? NT? OS/2? And do you mean 'compiles under', produces
programs that run under? 'uses features of', 'has library allowing
convenient access to API of'?
  Alsys, Rational (previously Meridian), and RR sell PC compilers
and GNAT is a Free Software Foundation partial Ada 95 compiler.
From the FAQ:
  Alsys:Pat Michalowski at 619/457-2700 for more info.
  Meridian:Gary Newman,800/653-2522, 703/318-5810.
  RR:Randall Brukardt, 800-PC-ADA4U, 608/251-3133, rbrukardt@bix.com
Gnat is ftp-able from pub/gnat at cs.nyu.edu
Aetech (whose number I don't find handy) also sells a PC compiler,
targeted, I believe, at Windows.

Re: Ada compiler for PC?

[Tom Moran]

>I am curious if there is a latest version of an Ada compiler
>anywhere for download, but I more interested in it being for a Win98 
>platform rather than the for-free aspect of it. 
Several companies (www.aonix.com, www.gnat.com, www.rrsoftware.com
and, I think, Greenhils, DDC, and Idaho Engineering Labs) offer Window
Ada compilers  If you want one for the price of a download (but
obviously without free support) look at Aonix (size limits on the
downloadable compiler) or Gnat.
  If "latest version" means in-house in alpha testing, you won't find
it.  The latest version that's publicly available for download is the
version that's publicly available for download.

Re: Ada compiler for PC?

[Steven Hovater]

Rational's  ApexNT product is pretty kick-a**, I would add.


Cheers,
Steve

Michael Feher wrote:

> Hello -
>    I know this is probably in a FAQ somewhere but here it goes...
>
>    I am curious if there is a latest version of an Ada compiler
> anywhere for download, but I more interested in it being for a Win98
> platform rather than the for-free aspect of it.  If anyone has
> any information for me, please respond to this newsgroup or
> e-mail me.
>
>    Thanks in advance -
>    Mike

--
Steven
Hovater
svh@rational.com
Software Engineering
Consultant
Phone/fax:781-676-2565/2500
Rational
Software
Pager: 888-906-2209
83 Hartwell Ave, Lexington,
MA                                             Amateur radio: AA1YH

Ada compiler for PC?

[Michael Feher]

Hello -
   I know this is probably in a FAQ somewhere but here it goes...

   I am curious if there is a latest version of an Ada compiler
anywhere for download, but I more interested in it being for a Win98 
platform rather than the for-free aspect of it.  If anyone has 
any information for me, please respond to this newsgroup or
e-mail me.

   Thanks in advance -
   Mike

Re: Ada compiler for PC?

[Wilhelm Spickermann]

Michael Feher wrote:
> 
> Hello -
>    I know this is probably in a FAQ somewhere but here it goes...
> 
>    I am curious if there is a latest version of an Ada compiler
> anywhere for download, but I more interested in it being for a Win98
> platform rather than the for-free aspect of it.  If anyone has
> any information for me, please respond to this newsgroup or
> e-mail me.
> 
>    Thanks in advance -
>    Mike
Hello -

Have a look at http://stad.dsl.nl/~jvandyk/compilers.html

Wilhelm Spickermann

Re: Ada compiler for PC?

[Tom Moran]

>Several companies ... offer Windows Ada compilers
Egad, I left off Rational (www.rational.com)

GNAT versions ( was :Ada compiler for PC?)

[SpamSpamSpam]

> When someone asks
> for the latest version of PowerPoint, they are not asking
> for the development version that some team in Redmond is
> using to test out their latest additions and fixes ...

Yes indeed, but then Power Point isn't GPL software ....

Glen.

Re: Ada compiler for PC?

[Larry Kilgallen]

In article <3718d384.254178@news.pacbell.net>, tmoran@bix.com (Tom Moran) writes:
>>I am curious if there is a latest version of an Ada compiler
>>anywhere for download, but I more interested in it being for a Win98 
>>platform rather than the for-free aspect of it. 
> Several companies (www.aonix.com, www.gnat.com, www.rrsoftware.com
> and, I think, Greenhils, DDC, and Idaho Engineering Labs) offer Window
> Ada compilers  If you want one for the price of a download (but
> obviously without free support) look at Aonix (size limits on the
> downloadable compiler) or Gnat.
>   If "latest version" means in-house in alpha testing, you won't find
> it.  The latest version that's publicly available for download is the
> version that's publicly available for download.

On the other hand, I think it possible that "latest" meant "conforming
to the most recent Ada standard".  Since that standard was completed
in 1995, it is the one addressed by all of those compilers.

Larry Kilgallen

Re: Ada compiler for PC?

[Robert Dewar]

In article <1999Apr20.073527.1@eisner>,
  Kilgallen@eisner.decus.org.nospam wrote:
> On the other hand, I think it possible that "latest"
> meant "conforming to the most recent Ada standard".
> Since that standard was completed
> in 1995, it is the one addressed by all of those
> compilers.


Indeed! I found Tom's reply very odd. When someone asks
for the latest version of PowerPoint, they are not asking
for the development version that some team in Redmond is
using to test out their latest additions and fixes ...

For the record,

The latest public version of GNAT is 3.11p

The latest version of GNAT Professional is 3.11b2

We are testing version 3.12 technology, and early beta
versions are being tested by selected customers. We are
hoping that the 3.12a GNAT Professional release will be
ready for beta testing within a couple of weeks. If this
testing goes smoothly, then a final release of GNAT
Professional 3.12 will follow, followed by a release
of public version 3.12p.

Robert Dewar

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: Ada compiler for PC?

[Tom Moran]

>When someone asks
>for the latest version of PowerPoint, they are not asking
>for the development version that some team in Redmond is
>using to test out their latest additions and fixes
 But the original asked:
> if there is a latest version of an Ada compiler
>anywhere for download
Of the finite set of versions available for download , one is
necessarily the latest of that set.  Presumably the Redmond
development version of PowerPoint, while later than any other, is not
in the set of available versions.  If it is, then it's probably the
latest in that set.  Similarly with Gnat - the daily development
version is not available for (public) download so it is not a member
of the set of "available for download" versions and thus is not the
latest member of that set.  Last time I looked,  3.11p was both
"available" and "the latest available", for download.

Re: GNAT versions ( was :Ada compiler for PC?)

[Larry Kilgallen]

In article <371DF7CE.C7D7C1F@spam.com>, Glen <spam@spam.com> writes:
> Robert Dewar wrote:
> 
>> In article <371C99DD.F15ADC4B@spam.com>,
>>   SpamSpamSpam <spam@spam.com> wrote:
>> > > When someone asks
>> > > for the latest version of PowerPoint, they are not
>> > > asking
>> > > for the development version that some team in Redmond
>> > > is using to test out their latest additions and fixes
>> ...
>> >
>> > Yes indeed, but then Power Point isn't GPL software ....
>>
>> So what? Why should the license under which sotware is
>> distributed affect anything with respect to proper quality
>> control ...
> 
> And just what affect does releasing "Betas" to the public haveon ACTs
> quality control ?

Presumably it would increase the number of posts on this issue
in comp.lang.ada, although that seems hard to believe.

Please take these endless disputes (as distinguished from inquiries)
to a software licensing forum and keep them out of comp.lang.ada.

Larry Kilgallen

Re: GNAT versions ( was :Ada compiler for PC?)

[Robert Dewar]

In article <371DF7CE.C7D7C1F@spam.com>,
  spamwithchipsplease@spam.com wrote:

> And just what affect does releasing "Betas" to the public
> have on ACTs quality control ?

You are confused, we are not talking beta releases here
but current development versions, which are often quite
unstable, and definitely unsuitable for release to anyone!

> And just how long where glib2 public users kept waiting
> for the "public" release of a GPL work available to
> paying support customers ? glitch free ?

We released this at the appropriate time in our release
cycle. Actually the procedures required for modifying the
sources were published on chat@gnat.com long before the
release.

> You didn't decide to use the GPL, the code came with the
> GPL,

Wrong, we wrote the Ada compiler, and we most definitely
decided to use the GPL. It indeed was me personally who
encouraged Chris to make this requirement! Of course the
decision to use gcc for the backend meant that it was
natural in any case to GPL the code, but the government
contract did not require the use of gcc, that was a
decision that I and my colleages at NYU took. So, you
are quite wrong here, we *did* decide to use the GPL, and
the code for the GNAT front end did not "come" with the
GPL. Indeed the starting point for the code, Ada/Ed, was
most definitely NOT GPL'ed!

> How does it limit my use, except for trying to make it
> proprietary ?

It actually has quite a lot of restrictions, you should
read it carefully. For example, if you manufacture a chip
which you are keeping secret for now, e.g. the Merced, and
you then get a company to build you a gcc based compiler,
then under the terms of the GPL, you cannot distribute
this to your customers under non-disclosure agreements.

It's actually quite interesting. One company can develop
a version of GNAT or GCC under wraps without violating the
GPL, but there is no way at all for two or more companies
to cooperate in such an endeavor.

You also can't distribute modified versions except under
restrictive conditions (source must be made available at
reasonable cost).

As I say, a limited license for use, that's what the GPL
is, and in that respect it is like any other software
license, it is just that its limitations are less severe!

> Microsoft prohibit me asking a fellow for a copy of the
> latest version of power point, the GPL allows anyone the
> right to give me or anyone else the latest version. This
> "not till its glitch free" is crap,

Well fine, if you think it is crap, you are welcome to
write your own software and do what you think proper with
it.

> you restrict a GPL work and all that might be learnt from
> such betas and an open testing, bug reporting, knowledge
> base, to protect your only asset,

Nope, our procedures have nothing to do with asset
protection, and everything to do with quality control.

> which isn't the  compiler because you don't own it

Well actually ownership is shared. The original version
of GNAT was owned by NYU, but under the terms of the
government contract, NYU was required to assign the
ownership to the FSF, so the basic compiler is owned
by the FSF. Subsequent modifications to these parts of
the system by ACT are also assigned to the FSF.

Part of the runtime library is owned by Florida State
University. This is the tasking runtime that is used in
the current versions of GNAT.

Other parts of the system are owned by ACT. Currently all
parts of the system are under the GPL or the GMG (someone
used these initials for the GNAT-modified-GPL, so I will
adopt that usage). The FSF is committed by the assignment
instrument to keep its part permanently GPL'ed. FSU and
ACT could theoretically take the parts they have
copyrighted and make new versions with different licesnses,
but this is unlikely to happen since

a) neither FSU nor ACT have any intention of deviating from
the commitment to make everything released under the GPL or
GMG. Unlike some other "free software" companies, ACT
remains 100% committed to the free software notion.

b) in any case the old versions are released under the GPL.
Suppose that FSU suddenly decided that all future versions
of the FSU components would be released under a restrictive
license in an attempt to make money by selling them. Well
in this case, ACT would simply take the last GPL'ed version
and continue to develop it, ignoring the FSU work. I don't
see that happening, as stated in a), but if it did it would
not have any significant effect.

By the way, from the above, I vaguely get the feeling that
you think that releasing something under the GPL is somehow
incompatible with ownership. That's not at all true from a
legal point of view. Indeed the GPL is effective ONLY
because the copyright holder intends to enforce their
copyright rigorously.

For example, if we found someone distributing GNAT in
a manner incompatible with the GPL, then we would inform
the FSF, and both the FSF and ACT would take action. First
we would advise the party to cease and desist, and if we
did not get satisfaction, we would file a breach of
copyright suit, in the same way that Micrsoft would if
their copyright were violated.

Robert Dewar
Ada Core Technologies

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GNAT versions ( was :Ada compiler for PC?)

[root]

Larry Kilgallen wrote:

> Please take these endless disputes (as distinguished from inquiries)
> to a software licensing forum and keep them out of comp.lang.ada.
> 

Note that I did cross posted to gnu.misc.discuss, so if anyone there
is to follow up please keep it in g.m.d.

I do sense that I am the ONLY C.L.A reader who disagrees with ACT policy
of GPL  versioning and release, so I shall cease in picking up when
Robert 
posts regards this subject ( even when he offers www.opensource.com
instead of www.gnu.org for a definition of "free software". )

Glen.

Re: GNAT versions ( was :Ada compiler for PC?)

[Robert Dewar]

In article <371C99DD.F15ADC4B@spam.com>,
  SpamSpamSpam <spam@spam.com> wrote:
> > When someone asks
> > for the latest version of PowerPoint, they are not
> > asking
> > for the development version that some team in Redmond
> > is using to test out their latest additions and fixes
...
>
> Yes indeed, but then Power Point isn't GPL software ....

So what? Why should the license under which sotware is
distributed affect anything with respect to proper quality
control ... or are you under the illusion that if I decide
to use the GPL as a license it forces me to distribute
untested stuff? If so, this is a misprehension. The GPL
is a license for limited use of software, certainly more
generous than the corresponding license for limited use
used by Microsoft, but again, so what?


-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GNAT versions ( was :Ada compiler for PC?)

[Glen]

Robert Dewar wrote:

> In article <371C99DD.F15ADC4B@spam.com>,
>   SpamSpamSpam <spam@spam.com> wrote:
> > > When someone asks
> > > for the latest version of PowerPoint, they are not
> > > asking
> > > for the development version that some team in Redmond
> > > is using to test out their latest additions and fixes
> ...
> >
> > Yes indeed, but then Power Point isn't GPL software ....
>
> So what? Why should the license under which sotware is
> distributed affect anything with respect to proper quality
> control ...

And just what affect does releasing "Betas" to the public haveon ACTs
quality control ? does your source code suddenly
become degraded or "infected" by the efforts of others ? Are your
paying customers so stupid as to take copies from unoffical sources ?
No.

And what about non-beta "professional" versions of GPL software
which are distributed to customers  long before the glitch free version
of the "public" release appears? how does restricting their distribution
affect quality ?

And just how long where glib2 public users kept waiting for the "public"
release of a GPL work available to paying support customers ? glitch free ?

> or are you under the illusion that if I decide
> to use the GPL as a license it forces me to distribute
> untested stuff?

You didn't decide to use the GPL, the code came with the GPL, anyadditions
you made  (and chose to release) HAD to be GPL.

> If so, this is a misprehension.

No, its your misrepresentation, just a spin  todivert from the fact that
ACT  cannot release a non GPL
GNAT.

> The GPL
> is a license for limited use of software,

How does it limit my use, except for trying to make it proprietary ?

> certainly more
> generous than the corresponding license for limited use
> used by Microsoft, but again, so what?

Microsoft prohibit me asking a fellow for a copy of the latest
version of power point, the GPL allows anyone the right to give
me or anyone else the latest version. This "not till its glitch free" is
crap, you restrict a GPL work and all that might be learnt from such
betas and an open testing, bug reporting, knowledge base,
to protect your only asset, which isn't the compiler because you
don't own it ( please don't talk about the version you keep under
your bed at ACT, which okay you do own ) its the knowledge.

Re: GNAT versions ( was :Ada compiler for PC?)

[Glen]

Robert Dewar wrote:
> 
> In article <371E2F2A.662C8F4F@spam.com>,
>   root <roffey@shagbadger.demon.co.uk> wrote:
> > Robert posts regards this subject ( even when he offers
> > www.opensource.com instead of www.gnu.org for a
> > definition of "free software". )
> 
> Please note that opensource is an appropriate source of
> information on open source software, NOT on free software.
> I have not ever suggested finding out about free software
> from open source software.
> 
http://www.dejanews.com/[ST_rn=ps]/getdoc.xp?AN=460428521

prick.

To my fellow C.L.A readers, sorry last and final no matter
what, I feel like Al Pacino in Godfather III "everytime I 
try to get out they draw me back in"

Could someone on GNU.MISC.DISCUSS please read the threads,
and pick up, this is the reality of commercial development
of GPL - proprietory versions !!

http://www.dejanews.com/[ST_rn=ps]/getdoc.xp?AN=451144738

Re: GNAT versions ( was :Ada compiler for PC?)

[Russell Senior]

>>>>> "Robert" == Robert Dewar <robert_dewar@my-dejanews.com> writes:

Robert> Indeed the whole point of the GPL is to avoid the
Robert> disadvantages of the proprietary model of software.

Perhaps you could explain a little bit more about how it is in the Ada
Community's interest to not share the GPL'd commercial and wavefront
releases of GNAT.  Not being a member of that community, it is the
part I don't understand.

Thanks.

-- 
Russell Senior
seniorr@teleport.com

Re: GNAT versions ( was :Ada compiler for PC?)

[dennison]

In article <371E2F2A.662C8F4F@spam.com>,
  root <roffey@shagbadger.demon.co.uk> wrote:

> I do sense that I am the ONLY C.L.A reader who disagrees with ACT policy
> of GPL  versioning and release, so I shall cease in picking up when
> Robert
> posts regards this subject ( even when he offers www.opensource.com
> instead of www.gnu.org for a definition of "free software". )

Actually, you are not. But it is generally accepted that they are well within
their rights to be doing it the way they are doing it. One might try to
*convince* them to change, but ACT has made their position on the matter
quite clear. It has been discussed to death already, and ACT remains unswayed
by the argments given. Unless there's some new compelling rationale that
no-one has yet thought of, it doesn't seem to be a productive discussion to
be engaging in.

--
T.E.D.

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GNAT versions ( was :Ada compiler for PC?)

[Robert Dewar]

In article <371E2F2A.662C8F4F@spam.com>,
  root <roffey@shagbadger.demon.co.uk> wrote:
> Robert posts regards this subject ( even when he offers
> www.opensource.com instead of www.gnu.org for a
> definition of "free software". )

Please note that opensource is an appropriate source of
information on open source software, NOT on free software.
I have not ever suggested finding out about free software
from open source software.

This is perhaps a good opportunity to point out the
distinction. The emphasis in the open source community
is on quality achieved by making sources openly available.

The emphasis in free software is on providing freedom to
the user of the software.

They are not the same thing, although there is some
overlap. The GPL for instance qualifies software as being
open source. However, the APL, recently accepted by the
open source organization, is definitely NOT free software,
since it places definite limitations on what you can do
with the software you receive under this license that are
incompatible with the aims of free software.

For more information on this important distinction, you
should indeed visit www.gnu.org.

GNAT remains free software, and in general free software
will always meet the requirements for open source software,
but it is very important to understand that the reverse is
NOT the case.

Robert Dewar

P.S as for Glen being unhappy with the product we produce,
I guess I cannot say more than that it is impossible to
please everyone with a given product. This is a free and
open market place. If Glen is unhappy with the GNAT
product, I guess he will simply have to shop elsewhere
for an Ada compiler. GNAT is not the only shop in town!

It will remain the case that ACT is, as befits the Ada
world, very much focussed on quality. Sometimes there is
a conflict between ensuring quality and getting the latest
and greatest version out right away without adequate
testing. ACT chooses to resolve this conflict with an
emphasis on careful QA for all our released products.

This does not mean that all our products will be 100%
completely problem free. We work towards that goal
certainly, but if that was the requirement I am afraid
people would get even more unhappy about delayed releases
:-)

Now, as is appropriate to this discussion, a short update
on the 3.12 status.

We have closed further development on 3.12 and are now
stabilizing the release. It will have many nice new
features, as has been consistently true of new GNAT
releases, and we will try to get it into good shape for
release as soon as possible.

Robert Dewar
Ada Core Technologies

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GNAT versions ( was :Ada compiler for PC?)

[Robert Dewar]

In article <371F8B02.9BD7045F@spam.com>,
  Glen <SpamSpam@spam.com> wrote:
> Could someone on GNU.MISC.DISCUSS please read the
> threads, and pick up, this is the reality of commercial
> development of GPL - proprietory versions !!

The GPL and the notion of proprietary software are
fundamentally incompatible. No versions of GNAT for
example are, or could be proprietary.

You must be using proprietary in an exceedingly unusual
sense, different from how most people use the term.

Indeed the whole point of the GPL is to avoid the
disadvantages of the proprietary model of software.

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GNAT versions ( was :Ada compiler for PC?)

[Pascal Obry]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 607 bytes --]


Russell Senior a �crit dans le message <86lnfjx5rx.fsf@coulee.tdb.com>...
>>>>>> "Robert" == Robert Dewar <robert_dewar@my-dejanews.com> writes:
>
>Robert> Indeed the whole point of the GPL is to avoid the
>Robert> disadvantages of the proprietary model of software.
>
>Perhaps you could explain a little bit more about how it is in the Ada
>Community's interest to not share the GPL'd commercial and wavefront
>releases of GNAT.  Not being a member of that community, it is the
>part I don't understand.
>


That's not a problem since members of the Ada community understand
that part :-)

Pascal.

Re: GNAT versions ( was :Ada compiler for PC?)

[bourguet]

In article <86lnfjx5rx.fsf@coulee.tdb.com>,
  Russell Senior <seniorr@teleport.com> wrote:
> >>>>> "Robert" == Robert Dewar <robert_dewar@my-dejanews.com> writes:
>
> Robert> Indeed the whole point of the GPL is to avoid the
> Robert> disadvantages of the proprietary model of software.
>
> Perhaps you could explain a little bit more about how it is in the Ada
> Community's interest to not share the GPL'd commercial and wavefront
> releases of GNAT.  Not being a member of that community, it is the
> part I don't understand.

First wavefront version of gnat are not release but more like snapshot,
and commercial release usually are quickly followed by a public one of the
same code.

About the public availability you may have another point of view than the
one of Robert Dewar, but his POV is quite argumented, consistant and in
nothing contrary to the GPL. The beta testers of emacs or gcc are perhaps
not the customers of a compagny, but snapshot and prerelease of these
programs are not very much more easily available than wavefront or commercial
release of gnats, and public release no less distant that public release of
gnat.

-- Jean-Marc
PS: Followup set to gnu.misc.discuss, this subject has been too much handled
on cla and nothing new will probably come out this time.

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GNAT versions ( was :Ada compiler for PC?)

[dennison]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 1302 bytes --]

In article <7fp5na$loj$1@cf01.edf.fr>,
  "Pascal Obry" <p.obry@der.edf.fr> wrote:
>
> Russell Senior a �crit dans le message <86lnfjx5rx.fsf@coulee.tdb.com>...
> >>>>>> "Robert" == Robert Dewar <robert_dewar@my-dejanews.com> writes:
> >
> >Robert> Indeed the whole point of the GPL is to avoid the
> >Robert> disadvantages of the proprietary model of software.
> >
> >Perhaps you could explain a little bit more about how it is in the Ada
> >Community's interest to not share the GPL'd commercial and wavefront
> >releases of GNAT.  Not being a member of that community, it is the
> >part I don't understand.
> >
>
> That's not a problem since members of the Ada community understand
> that part :-)

Well, perhaps some of us do...

But my personal interest or my personal agreement as a "member of the
community" isn't the issue. The issue is if this is all kosher by the GPL,
which it pretty clearly is.

Given several cases of beer and an evening to blow, I'm sure we could all as
a group come up with dozens of "better" development models for Gnat. But who
really cares? ACT certainly doesn't have to, and I don't even think I do...

--
T.E.D.

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GNAT versions ( was :Ada compiler for PC?)

[Samuel Mize]

In comp.lang.ada Russell Senior <seniorr@teleport.com> wrote:
>>>>>> "Robert" == Robert Dewar <robert_dewar@my-dejanews.com> writes:
> 
> Robert> Indeed the whole point of the GPL is to avoid the
> Robert> disadvantages of the proprietary model of software.
> 
> Perhaps you could explain a little bit more about how it is in the Ada
> Community's interest to not share the GPL'd commercial and wavefront
> releases of GNAT.  Not being a member of that community, it is the
> part I don't understand.

The commercial one IS shared.  The commercial customer can always get
all the source code, and is allowed to give out copies.

Of course, if you didn't get it via a formal distribution from ACT
they don't guarantee that it's uncorrupted, and if you aren't paying
money to ACT they don't promise to support you.

The public version is the most recent stable commercial version,
"marked" as not formally distributed and therefore unsupported.

"Wavefronts" are very early versions of things which, once fully
debugged, will go into some commercial version and be released.

A customer who gets a "wavefront" is, IIRC, legally allowed to give it
out.  But the wavefront recipients know that it's a preliminary work
in progress, and not yet ready for wide distribution.  Most don't
distribute it.

If you have evidence that threats are used to keep someone from giving
out a wavefront, you may have evidence of action that goes against the
intent of the GPL.  But bear in mind that getting a wavefront is a
favor, not a right.  Most of us don't consider it a "threat" for ACT
to be less inclined to do a favor for someone who has annoyed them,
e.g. by giving out a prior wavefront.

If the "wavefront" mechanism were being used to withhold useful or
important changes from the community at large, that would not be nice,
but (again, IIRC) it still wouldn't violate the GPL.  The GPL is meant
to ensure that the source code will be available to the CUSTOMER who
gets a binary, not necessarily to the whole planet.

And, I don't know of any real allegations that something important was
being withheld for several releases.  I've just heard complaints based
in a desire to see and comment on work in progress, and the occasional
vague suspicion that wavefronts have "the good stuff."

Bear in mind that many compiler vendors will refuse to give out a
preliminary version, even to a "supported" customer.  You just get a
promise that "we'll fix that in a later release."  I find it amusing
that a company who goes further to make its code available is
therefore accused of withholding it.

Best,
Sam Mize

-- 
Samuel Mize -- smize@imagin.net (home email) -- Team Ada
Fight Spam: see http://www.cauce.org/ \\\ Smert Spamonam

Re: GNAT versions ( was :Ada compiler for PC?)

[Russell Senior]

>>>>> "Samuel" == Samuel Mize <smize@imagin.net> writes:

>>>>> "Robert" == Robert Dewar <robert_dewar@my-dejanews.com> writes:

Robert> Indeed the whole point of the GPL is to avoid the
Robert> disadvantages of the proprietary model of software.

Russell> Perhaps you could explain a little bit more about how it is
Russell> in the Ada Community's interest to not share the GPL'd
Russell> commercial and wavefront releases of GNAT.  Not being a
Russell> member of that community, it is the part I don't understand.

Samuel> The commercial one IS shared.  The commercial customer can
Samuel> always get all the source code, and is allowed to give out
Samuel> copies.

Thanks for the review.  I am not accusing anyone of anything.  I am
asking a question.  Of the three replies to my question that I saw
this morning, none of them addressed the question.

It has been noted that all of the GNAT releases (or snapshots or
whatever you'd like to call them) are GPL'd and that recipients are
legally entitled to share these with others.  It has further been
noted that recipients of the `non-public' releases tend not to share
them with others, despite their legal right to do so.  I am curious to
know what goes into their calculus for deciding not to.  That is all.

Any attempts to help educate me on this question would be most
welcome.  Thanks.

-- 
Russell Senior
seniorr@teleport.com

Re: GNAT versions ( was :Ada compiler for PC?)

[Robert Dewar]

In article <7fq3mh$16m1@news1.newsguy.com>,
  Samuel Mize <smize@imagin.net> wrote:

> I've just heard complaints based
> in a desire to see and comment on work in progress, and
> the occasional
> vague suspicion that wavefronts have "the good stuff."

Well some wavefronts have "the good stuff", but
unfortunately, from time to time, some of them, especially
the ones we do NOT send to anyone at all, have "the bad
stuff".

Asking for wavefronts to be widely distributed is sort of
like asking for drugs to be made available immediately
to the public before they are tested. It might save some
lives, but also may kill some people!

It's our job to separate the good from the bad, and make
sure that only the good stuff gets to the final releases.
We are getting pretty near that point with 3.12 now, and so
pretty soon the Ada community, both those using the
commercial version of GNAT Professional, and those using
the public version, will be able to get their hands on
all the good stuff.

I will post announcements here on CLA as the 3.12 release
process progresses. It will go through the following
stages:

1. Preliminary beta release for selected targets to
selected users.

2. General beta release of GNAT Professional available
to all our customers, available on almost all targets.

3. Official release of GNAT Professional, available on
all supported targets.

4. Release of the public version on all targets.

One thing that we like to do if we can is to distribute
for all targets at about the same time (we get yelled at
loudly on CLA and elsewhere if we don't release for all
targets at about the same time, since that understandably
frustrates people -- yes yes I know the OS/2 users are
frustrated right now for example, and this is a situation
that we try to avoid). So we like to wait till kinks on
all targets are ironed out before distributing the public
version.

But everything is going smoothly, there are significant
functionality improvements, performance improvements,
new tools, and many bugs fixed in 3.12.

For those interested in learning more about our release
procedures and policies, they have been discussed at length
on CLA in the past, so a scan of the Dejanews archives will
undoubtedly tell you all that you want to know, and more
on this issue.

Robert Dewar
Ada Core Technologies

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GNAT versions ( was :Ada compiler for PC?)

[Robert Dewar]

In article <86n1zz43dt.fsf@coulee.tdb.com>,
  Russell Senior <seniorr@teleport.com> wrote:

> It has further been
> noted that recipients of the `non-public' releases tend
> not to share them with others, despite their legal right
> to do so.  I am curious to
> know what goes into their calculus for deciding not to.
> That is all.

People have many motives for their actions. Just because
something is legal does not mean it is a good idea to do
it. Probably in practice, when it comes to wavefronts,
people realize that the are only useful in the specific
case where they are necessary to solve a particular
bug. Note that we do not even make wavefronts available
to all our customers, we only make them available in
specific cases if in our judgment it is the only way
to work around a specific problem. Distributing these
work-in-progress snap-shots can certainly cause a lot
of confusion. As you know, we are of the opinion that
it is undesirable to the Ada community as a whole to
distribute these widely. Sure there are a few hackers
and enthusiasts who would like to fiddle with the latest
and greatest, but we have to consider the interests of
the community as a whole, and I guess the fact that the
wavefronts do not get redistributed is an indication that
our customers agree with our judgment here. Indeed most
of our customers would prefer NOT to get wavefronts, and
instead work around problems with our help in the official
releases, and that seems perfectly reasonable.

Robert Dewar
Ada Core Technologies

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GNAT versions ( was :Ada compiler for PC?)

[Robert Dewar]

In article <7fqb0h$7r7$1@nnrp1.dejanews.com>,
  dennison@telepath.com wrote:

> Given several cases of beer and an evening to blow, I'm
> sure we could all as a group come up with dozens of
> "better" development models for Gnat. But who
> really cares? ACT certainly doesn't have to, and I don't
> even think I do...

How about using that case of beer and the evening to
discuss how to persuade others in the Ada vendor community
(compilers .. Greenhills, Rational, Aonix, OCS ... etc
tools, Claw, etc etc) to look more closely at the free
softwaremodel for their products.

I quite understand that people get
exercised over the issue of whether they see the
latest feature in GNAT in June as opposed to July, but
what would really help the Ada community is if more Ada
suppliers would climb on the open software bandwagon,
preferably using proper free software licenses!

It's good that GNAT has persuaded some more free releases
(e.g. the crippled versions of the Aonix compiler, and
CLAW), but it would be *so* much better if tools like this
could genuinely be made openly available without any
limitations or "demo" versions.

Don't assume it can't happen. The open source movement is
gathering a lot of steam. When we started out on this track
6 years ago, we were pioneering, now we are in the
mainstream!

Robert Dewar
Ada Core Technologies

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GNAT versions ( was :Ada compiler for PC?)

[Samuel Mize]

In comp.lang.ada Russell Senior <seniorr@teleport.com> wrote:

> Thanks for the review.  I am not accusing anyone of anything.  I am
> asking a question.  Of the three replies to my question that I saw
> this morning, none of them addressed the question.

That's because there are some people who take an angry stance about
this and ride it into the dirt.  (I mean, agree with them or not,
but they go on and on and on...)  So people here are a little
twitchy.  I'm glad this isn't your intent.

> It has been noted that ...
> recipients of the `non-public' [GNAT] releases tend not to share
> them with others, despite their legal right to do so.  I am curious to
> know what goes into their calculus for deciding not to.  That is all.

There are two kind of non-public releases, "wavefronts" and
"commercial releases."

A wavefront is an in-development copy that may not be fully tested,
hasn't been regression tested, hasn't been through QA, and may be
quite unstable.  It's given to a customer who badly needs a new
capability, or perhaps the fix to a specific bug.

A commercial release is a formally released product that ACT has
stabilized, put through regression test, beta test and QA, and
generally believes is ready to be released as a commercial product.

ACT doesn't want wavefronts passed around widely, because they're a
snapshot from the middle of development, and they expect it to have
problems.  They don't want to be flooded with bug reports on code that
they weren't done with in the first place.

ACT's customers go along because they understand and accept ACT's
rationale.  They may also want to encourage ACT to give them other
wavefronts later.  Bear in mind, ACT has no obligation to do so.  They
are only obliged to release code when its binary is delivered to a
customer; and even then, IIRC, only to that customer.

ACT's customers seldom bother to distribute the commercial version,
because ACT will publish it soon as a public version anyway.  One
might be able to beg a copy from a buddy who's a supported customer
if, for instance, one were having trouble with a compiler bug that is
fixed in that release.  Nothing wrong with that.

Once the commercial version is fully stable, ACT cuts a new copy of
that release, which is the "public" version.  The only change from the
commercial version is the version number.  This will lag the
commercial version.  This is partly because this is non-paying work,
and partly so that any bugs which are found by early adopters can be
fixed for the public version.  (Again, to avoid a flood of redundant
bug reports when the thundering horde of free users gets hold of the
product.)

So, where's the conflict coming from?

Some people would like to see wavefronts issued publicly and
periodically -- even frequently.  I believe some want to feel they
have the very "cutting edge technology" compiler, some may have wanted
a more-recent version in the hope that a specific bug was fixed, and
some just want to look into the development process.  These people are
not in a position to get wavefronts from ACT under ACT's current
policies, and they get frustrated and vocal about it.

I feel that there are fair arguments on both sides.  I'm biased toward
the ACT position in this specific case, and that's probably clear from
my writing.  GNAT is large and complex, and it's unlikely IMHO that
adding a few thousand more cooks will improve the broth.

On the other hand, I certainly understand the frustration of people
who would like to be more involved in its evolution.  I hope they get
together, set up some kind of outside-ACT development tree, and
coordinate with ACT.  One of the goals of GNAT was to encourage
experimentation with new ideas and extensions to Ada, and I feel this
would encourage work in that area.

But on the other other hand, this isn't ACT's job.  People who want to
have a parallel-and-coordinated development effort must set it up on
their own.  IMHO.

I hope you've found this helpful.

Best,
Sam Mize

-- 
Samuel Mize -- smize@imagin.net (home email) -- Team Ada
Fight Spam: see http://www.cauce.org/ \\\ Smert Spamonam

Re: GNAT versions ( was :Ada compiler for PC?)

[Larry Kilgallen]

Reply-To: Kilgallen@eisner.decus.org.nospam
Organization: LJK Software
Lines: 26

In article <7fqld6$htu$1@nnrp1.dejanews.com>, Robert Dewar <robert_dewar@my-dejanews.com> writes:
> In article <86n1zz43dt.fsf@coulee.tdb.com>,
>   Russell Senior <seniorr@teleport.com> wrote:
> 
>> It has further been
>> noted that recipients of the `non-public' releases tend
>> not to share them with others, despite their legal right
>> to do so.  I am curious to
>> know what goes into their calculus for deciding not to.
>> That is all.
> 
> People have many motives for their actions. Just because
> something is legal does not mean it is a good idea to do
> it. Probably in practice, when it comes to wavefronts,
> people realize that the are only useful in the specific
> case where they are necessary to solve a particular
> bug. Note that we do not even make wavefronts available
> to all our customers, we only make them available in
> specific cases if in our judgment it is the only way
> to work around a specific problem.

It seems unlikely that someone with "a specific problem"
would be at a point in their own work cycle where they
have time to get into the software distribution business.

Larry Kilgallen

Re: GNAT versions ( was :Ada compiler for PC?)

[Christopher Browne]

On Fri, 23 Apr 1999 08:58:44 +0200, Pascal Obry <p.obry@der.edf.fr> wrote:
>
>Russell Senior a �crit dans le message <86lnfjx5rx.fsf@coulee.tdb.com>...
>>>>>>> "Robert" == Robert Dewar <robert_dewar@my-dejanews.com> writes:
>>
>>Robert> Indeed the whole point of the GPL is to avoid the
>>Robert> disadvantages of the proprietary model of software.
>>
>>Perhaps you could explain a little bit more about how it is in the Ada
>>Community's interest to not share the GPL'd commercial and wavefront
>>releases of GNAT.  Not being a member of that community, it is the
>>part I don't understand.
>
>That's not a problem since members of the Ada community understand
>that part :-)

Hmm...  Sounds to me like a situation where there start to be
entertaining definitions of what "The Ada community" is.

"A *true* member of the Ada community is one that properly understands
the release strategy and agrees with it.  Anyone who doesn't
understand obviously isn't part of the community..."
-- 
Fatal Error: Found [MS-Windows] System -> Repartitioning Disk for Linux...
(By cbbrowne@hex.net, Christopher Browne)
cbbrowne@ntlug.org- <http://www.ntlug.org/~cbbrowne/languages.html>

Re: GNAT versions ( was :Ada compiler for PC?)

[Fergus Henderson]

kilgallen@eisner.decus.org (Larry Kilgallen) writes:

>It seems unlikely that someone with "a specific problem"
>would be at a point in their own work cycle where they
>have time to get into the software distribution business.

These days we have an amazing invention called "The Internet" which
makes distributing software very easy.

-- 
Fergus Henderson <fjh@cs.mu.oz.au>  |  "I have always known that the pursuit
WWW: <http://www.cs.mu.oz.au/~fjh>  |  of excellence is a lethal habit"
PGP: finger fjh@128.250.37.3        |     -- the last words of T. S. Garp.

Re: GNAT versions ( was :Ada compiler for PC?)

[Russ Allbery]

In gnu.misc.discuss, Russell Senior <seniorr@teleport.com> writes:

> It has been noted that all of the GNAT releases (or snapshots or
> whatever you'd like to call them) are GPL'd and that recipients are
> legally entitled to share these with others.  It has further been noted
> that recipients of the `non-public' releases tend not to share them with
> others, despite their legal right to do so.  I am curious to know what
> goes into their calculus for deciding not to.

Well, I can't speak for all the people who make that decision, but my
personal answer to such a question would be "simple politeness."  If
someone gives me something to test and tells me it's not ready for general
redistribution, I would tend to respect their wishes unless I had some
good reason not to.

-- 
Russ Allbery (rra@stanford.edu)         <URL:http://www.eyrie.org/~eagle/>

Re: GNAT versions ( was :Ada compiler for PC?)

[Ronald Cole]

Robert Dewar <robert_dewar@my-dejanews.com> writes:
> In article <371F8B02.9BD7045F@spam.com>,
>   Glen <SpamSpam@spam.com> wrote:
> > Could someone on GNU.MISC.DISCUSS please read the
> > threads, and pick up, this is the reality of commercial
> > development of GPL - proprietory versions !!
> 
> The GPL and the notion of proprietary software are
> fundamentally incompatible. No versions of GNAT for
> example are, or could be proprietary.
> 
> You must be using proprietary in an exceedingly unusual
> sense, different from how most people use the term.
> 
> Indeed the whole point of the GPL is to avoid the
> disadvantages of the proprietary model of software.

Glen, apparently, lacks the proper vocabulary to engage in a proper
discussion.  Perhaps I can help here...

The correct word is "hoarding".  And, yes, the GPL allows hoarding.
Stallman doesn't think that if you distribute GPL'd source code to
person A that you must also distribute it to person B if he asks.

Remember the joke about the difference between a slut and a bitch?
A slut puts out for everyone and a bitch puts out for everyone but you.
The GPL allows one to be a bitch.

-- 
Forte International, P.O. Box 1412, Ridgecrest, CA  93556-1412
Ronald Cole <ronald@forte-intl.com>      Phone: (760) 499-9142
President, CEO                             Fax: (760) 499-9152
My PGP fingerprint: 15 6E C7 91 5F AF 17 C4  24 93 CB 6B EB 38 B5 E5

Re: GNAT versions ( was :Ada compiler for PC?)

[Leslie Mikesell]

In article <7frqmj$bg6$1@mulga.cs.mu.oz.au>,
Fergus Henderson <fjh@cs.mu.oz.au> wrote:
>kilgallen@eisner.decus.org (Larry Kilgallen) writes:
>
>>It seems unlikely that someone with "a specific problem"
>>would be at a point in their own work cycle where they
>>have time to get into the software distribution business.
>
>These days we have an amazing invention called "The Internet" which
>makes distributing software very easy.

Yes, and wouldn't you expect everyone with the previous public
release to have that same 'specific problem'?  And for them,
the only way to get the fix is to become a commercial customer
at which time they magically become capable of dealing with
the other potential problems of the wavefront releases...


  Les Mikesell
    les@mcs.com

Re: EGCS & GNAT Was (Re: GNAT versions)

[Leslie Mikesell]

In article <1999Apr25.201259.1@eisner>,
Larry Kilgallen <kilgallen@eisner.decus.org> wrote:
>Reply-To: Kilgallen@eisner.decus.org.nospam
>Organization: LJK Software
>Lines: 16
>
>In article <37234DCA.A509D0D9@aasaa.ofe.org>, David Starner <dstarner98@aasaa.ofe.org> writes:
>
>> I still don't understand the great worries about GNAT beta versions
>> getting distributed.
>
>Neither do I, but people keep bringing this up in comp.lang.ada
>as a political football, without anyone coming up with a single
>example where there has been a problem with the current system
>of releasing only software that has received the fuller testing.

It is mostly bickering about the meaning of the GPL.  As in how
you can distribute even one bugfix copy of something with
a license that ensures the right to source and the right to
redistribute the source while witholding the source and insisting
that the recipient not redistribute?  One or the other doesn't
make sense.

>As an Ada fan, I would much prefer the versions of Ada available
>(from all sources) were as defect-free as possible, so those who
>are not already firmly in the Ada camp will not get disuaded by
>encountering defects.

If the old versions are perfect, why is there ongoing development?

  Les Mikesell
   les@mcs.com

Re: GNAT versions ( was :Ada compiler for PC?)

[Geoffrey KEATING]

bourguet@my-dejanews.com writes:

> First wavefront version of gnat are not release but more like snapshot,
> and commercial release usually are quickly followed by a public one of the
> same code.
> 
> About the public availability you may have another point of view than the
> one of Robert Dewar, but his POV is quite argumented, consistant and in
> nothing contrary to the GPL. The beta testers of emacs or gcc are perhaps
> not the customers of a compagny, but snapshot and prerelease of these
> programs are not very much more easily available than wavefront or commercial
> release of gnats, and public release no less distant that public release of
> gnat.

You mean, the wavefront and/or commercial releases of gnats are
available by anonymous FTP, and the current tree can be obtained by
anonymous CVS?  I'd be interested in more details, if so.

This is certainly true for egcs, and IIRC even for gcc there is
anonymous CVS access.

[Before anyone asks, you can get egcs snapshots from egcs.cygnus.com,
or any of its mirrors.  Releases are put in the same place.]

-- 
Geoff Keating <Geoff.Keating@anu.edu.au>

Re: GNAT versions ( was :Ada compiler for PC?)

[Robert Dewar]

In article <t8zp3xy1w7.fsf@discus.anu.edu.au>,
  Geoffrey KEATING <geoffk@discus.anu.edu.au> wrote:
> This is certainly true for egcs, and IIRC even for gcc
> there is anonymous CVS access.

But surely you do not think that the current EGCS snapshot
does not represent the internal Cygnus development tree
(actually you may think this, I quite often run into people
under this remarkable illusion!



-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GNAT versions ( was :Ada compiler for PC?)

[Robert Dewar]

In article <7fqjo2$1hmi@news1.newsguy.com>,
  Samuel Mize <smize@imagin.net> wrote:
> A wavefront is an in-development copy that may not be
> fully tested, hasn't been regression tested, hasn't been
> through QA, and may be quite unstable.

Actually that's *too* negative, we never let anything out
of ACT, even a wavefront, that has not been fully
regression tested against the ACVC suite, and against the
internal (at this stage VERY substantial ACT regression
suite). Indeed our normal internal procedure for ANY
internal fix, however small, is to run the ACT regression
suite before a checkin to make sure no regressions have
occurred.

Nevertheless, it is quite true that this does not represent
"fully tested", since part of full testing is field testing
at customer sites in a beta test program, which most
certainly does not happen for each wavefront. As for QA,
even a wavefront has been through our internal QA process,
including running the ACVC test suite, our ACT regression
suite, and the DEC test suite. However, once again, the
full QA process most certainly includes beta testing, and
this is only done on formal releases.

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GNAT versions ( was :Ada compiler for PC?)

[Robert Dewar]

In article <7ftfj4$vln$1@Jupiter.mcs.net>,
  les@MCS.COM (Leslie Mikesell) wrote:
> Yes, and wouldn't you expect everyone with the previous
> public release to have that same 'specific problem'?  And
> for them, the only way to get the fix is to become a
> commercial customer at which time they magically become
> capable of dealing with the other potential problems of
> the wavefront releases...

Well in fact it would NOT be good enough for them to become
a customer. We don't make wavefronts generally available
to our customers. Why not? Because in general it would
not be helpful. Yes, a development wavefront may fix
some problems, but until really adequate testing has
been done, which is generally not possible for wavefronts,
it may well introduce new problems, and new features that
have been added may intefere with your existing programs.

For example, if you have (unwisely but it happens) done
a WITH of one of the GNAT internal system implementation
units, then a wavefront may change that interface without
warning. Now this is something that you will have to deal
with in the major release cycle, but you probably do not
want to mess with this just on the possibility that the
wavefront may fix some problems you are not aware of.

A customer getting a wavefront is in the following
situation:

1. They are fixing something that they know is broken in
their code, and there is no other way to fix it. Most often
we can find a work around, and in that case we do NOT
provide a wavefront. But if there is no other way, then it
becomes worth the risk of moving to a new version of the
compiler without waiting for a new release.

2. When a customer gets a wavefront, they have direct
support from us, which means that if they do run into
problems, we can help them out.

I would *never* advise anyone to move to a wavefront unless
they were forced to. It seems a very bad idea to me to be
constantly switching compiler versions.

Obviously it would be possible to mitigate the risks of
wavefronts if we had the resources to develop two versions
simultaneously, the development version with all the new
features, and the old version with minimal bug fixes.
Indeed this is somewhat what Cygnus does with EGCS.
Obviously EGCS does not have the new development stuff
that Cygnus has internally (which is very closely held
until it is released), but they do spend significant
resources coordinating bug fixes with EGCS.

Perhaps in the future ACT will have more resources to do
this kind of dual development. It may happen at least to
some extent with the work of the GNAT/Linux group, at least
that is what we hope can be worked out.

But it is not a zero-cost proposition by any means. Note
for example that GDB has nothing analogous to the EGCS
process at the moment. Right now, Cygnus feels that they
are doing most of the GDB work anyway, so it does not make
sense. As time goes on and more people contribute directly
to GDB as seems likely to happen in the future, it will
make more sense to create a more open process for GDB.

Obviously if there were more players than ACT contributing
directly to the GNAT work, then the value of such
coordination would increase. This seems to be beginning
to happen with the GNAT/Linux work, one useful patch
recently propagated from the GNAT/Linux team to ACT, and
we have prepared a couple of critical Linux patches to go
in the opposite direction.

One problem of course is that only ACT has access to the
ACT regression suite. This is fundamental, it is mostly
proprietary code, that we promise to guard very closely.
Similarly, the DEC test suite is proprietary code to which
we have access, but most certainly not distribution rights.
Of course the ACVC suite exists publicly, and this is at
least a first level test of externally developed patches
and fixes.

Anyway we will see, The 3.12p release will provide a more
stable starting point for such interaction. Why more
stable? Because each release so far of GNAT, with rare
exceptions is more stable than the prior one. Why? Because
we work hard to make it so.

Robert Dewar
Ada Core Technologies

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GNAT versions ( was :Ada compiler for PC?)

[Russ Allbery]

In gnu.misc.discuss, Robert Dewar <robert_dewar@my-dejanews.com> writes:

> One problem of course is that only ACT has access to the ACT regression
> suite. This is fundamental, it is mostly proprietary code, that we
> promise to guard very closely.  Similarly, the DEC test suite is
> proprietary code to which we have access, but most certainly not
> distribution rights.

This is actually a really interesting issue that faces any developer of a
free software compiler.

Fundamental to compiler development of any sort is extensive regression
testing, since a bug in a compiler will not only cause the compiler to not
function correctly, but may also propagate breakage to everything compiled
with the compiler, sometimes in extremely subtle ways that are damnably
hard to detect.  Extensive testing is considerably more important even
than for other critical system utilities, and the need is probably
equalled only by system kernels and libc.

One of the best sorts of tests are things that used to be broken and are
now fixed, that can be tested with every change to make sure they don't
break again.  (That's the definition of regression testing, in fact.)  But
what a compiler breaks on is certain code constructs, and other people's
code is quite frequently not under the same license as the compiler
itself.  Therefore, if one actually uses one of the best sources of a test
suite for a compiler, one ends up with a test suite that's covered under a
bunch of conflicting licenses, including, quite frequently, proprietary
ones (since people still do write a lot of proprietary software), and
which almost certainly can't be easily distributed with the compiler.

Companies are usually willing to let their internal code be used for
testing the compiler that they actually use, but they're much less likely
to be willing to have that code redistributed.  But not adding a good test
to the test suite due to licensing issues isn't very palatable either, nor
is trying to rewrite the test.  (In fact, the latter is often not even
possible, since the obscure interactions that caused the bug in the first
place usually disappear in a rewrite.)

It makes a great deal of sense for there to be a "traditional" software
support company associated with a free compiler effort that can take bug
reports and test code from companies and other sources and perform
regression tests against the free source base using that information, even
if the test code can't be redistributed.

-- 
Russ Allbery (rra@stanford.edu)         <URL:http://www.eyrie.org/~eagle/>

Re: GNAT versions ( was :Ada compiler for PC?)

[Robert Dewar]

In article <yl676lnovr.fsf@windlord.stanford.edu>,
  Russ Allbery <rra@stanford.edu> wrote:

> In gnu.misc.discuss, Robert Dewar
> <robert_dewar@my-dejanews.com> writes:

<<many other good points snipped>>

> It makes a great deal of sense for there to be a
> "traditional" software support company associated with a
> free compiler effort that can take bug reports and test
> code from companies and other sources and perform
> regression tests against the free source base using that
> information, even if the test code can't be
> redistributed.

We certainly regard our test suite (now many millions of
lines long spreading out over nearly 6000 directories, and
close to a hundred thousand files) as a critical aid in
our quality control procedures. Other companies undoubtedly
have their own suites. Often this code is very fiercely
protected, and we routinely sign non-disclosure agreements
to get it.

A useful, but massive task, would be to sanitize this test
suite, by extracting the tests that are not subject to
non-disclosure (since this is the default for us, it would
take a large effort of contacting users etc), and creating
sanitized versions of bugs in proprietary code. That could
result in a useful freely available test. But I would
estimate that several person years of work is required
for this, and we just don't have the resources.

One very important point for us is that we also have
access to the Digital test suite from their extensively
used Ada 83 product (this includes basically all the bugs
that DEC ever ran into plus a number of really diabolical
complex purpose written tests [the ACT test suite also
includes hundreds of our own purpose-written tests,
generally when we add a new feature, we add a test to our
test suite].

In practice we find that the trio working together (the
ACVC tests, the ACT tests, and the DEC test suite) are
complementary, very often a new version of the compiler
passes 2 out of 3 of these tests!

We actually use the ACT regression suite very aggressively.
Our internal procedures allow anyone to checkin
modifications to one of the standard GNAT files, but *only*
after running the regression suite. This suite has been
run nearly ten thousand times in the last couple of years,
and we have found this a very effective way of controlling
development and making sure that regressions are not
introduced. Then each night we build on all targets, and
run the ACVC and DEC test suites.

In any case we are ahead in the Ada community in having
at least one freely available comprehensive test suite,
the ACVC tests, and anyone for example doing their own
fiddling with GNAT should most CERTAINLY be running the
ACVC tests as one good step to making sure they are not
introducing regressions.

Robert Dewar
Ada Core Technologies

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: EGCS & GNAT Was (Re: GNAT versions)

[David Starner]

Robert Dewar wrote:
> 
> In article <t8zp3xy1w7.fsf@discus.anu.edu.au>,
>   Geoffrey KEATING <geoffk@discus.anu.edu.au> wrote:
> > This is certainly true for egcs, and IIRC even for gcc
> > there is anonymous CVS access.
> 
> But surely you do not think that the current EGCS snapshot
> does not represent the internal Cygnus development tree
> (actually you may think this, I quite often run into people
> under this remarkable illusion!

But the internal Cygnus development tree doesn't represent the EGCS
development tree either. A great deal of development goes on in the EGCS
developement tree, and the internal Cygnus development tree contains
some stuff that is not suited for the EGCS tree (for example: a
enviroment variable that the compiler reads for command line options;
they deem this too error prone for EGCS). g77, if I understand
correctly, is completely developed in the outside tree.

I still don't understand the great worries about GNAT beta versions
getting distributed. I don't see problems arising from the EGCS
snapshots being distributed; if you download a snapshot you know what
you're getting. I don't see how big problems would arise from releasing
GNAT wavefront versions. It certainly doesn't compel ACT to distribute
beta version, though.

While it's on topic, since EGCS is now the official GCC tree, when will
there be a version of GNAT that builds on EGCS? It would be much more
convientent for those of use running EGCS.

Re: GNAT versions ( was :Ada compiler for PC?)

[Maciej Stachowiak]

Robert Dewar <robert_dewar@my-dejanews.com> writes:

> In article <7ftfj4$vln$1@Jupiter.mcs.net>,
>
> Obviously it would be possible to mitigate the risks of
> wavefronts if we had the resources to develop two versions
> simultaneously, the development version with all the new
> features, and the old version with minimal bug fixes.
> Indeed this is somewhat what Cygnus does with EGCS.
> Obviously EGCS does not have the new development stuff
> that Cygnus has internally (which is very closely held
> until it is released), but they do spend significant
> resources coordinating bug fixes with EGCS.
> 

This is not IMO a fair description of the EGCS project. EGCS contains
a lot more than bug fixes, it has a lot of new feature development as
well - new optimizations, C++ front-end work and new platform support,
among other things.

Cygnus has regularly delivered the code for major features to the EGCS
tree, and I suspect quite a few major features appeared in the EGCS
tree before they were shipped to Cygnus customers. In fact, EGCS began
as essentially a drop of the Cygnus internal development tree.

Of course, I am not a Cygnus employee so I can't really speak for
their proedures, but that is how it looks to me.

Also, even though Cygnus provides the infrastructure for EGCS and pays
a number of employees to work on it, EGCS is not really controlled by
Cygnus, but rather by a steering committee on which Cygnus employees
constitute a minority.

 - Maciej Stachowiak

Re: GNAT versions ( was :Ada compiler for PC?)

[Maciej Stachowiak]

Robert Dewar <robert_dewar@my-dejanews.com> writes:

> In article
> <m3g15ovkpk.fsf@207-172-255-37.s37.tnt2.sbo.ma.dialup.rcn.c
> om>,
>   Maciej Stachowiak <mstachow@alum.mit.edu> wrote:
> > Cygnus has regularly delivered the code for major
> > features to the EGCS tree, and I suspect quite a few
> > major features appeared in the EGCS tree before they were
> > shipped to Cygnus customers.
> 
> That would be an interesting suspicion to back up. 

One singinficant example would be the Java front end. That appeared in
the EGCS tree long before it was at all fit for use, so I doubt it got
shipped to customers first. From a rough examination of their web
pages, it is not clear whether they are selling it now.

> Yes, of course Cygnus drops developments to EGCS, no one
> is saying that they don't. 

In fact, a large reason they started it was to avoid maintaining large
internal code forks.

> It is just that, as in the case of GNAT, there is usually a lag
> between the development of new features and their dropping into
> EGCS.

There's a difference between lag time from development to drop to the
public tree and lag time between ship to paying customers and drop to
the public tree. Of course, Cygnus likely does both for different
things. 

I am not going to get into this flame war though, I am not even an Ada
user and I think anyone should have the right to use whatever release
policy they choose as long as they comply with relevant licenses.

 - Maciej

Re: GNAT versions ( was :Ada compiler for PC?)

[Robert Dewar]

In article
<m3g15ovkpk.fsf@207-172-255-37.s37.tnt2.sbo.ma.dialup.rcn.c
om>,
  Maciej Stachowiak <mstachow@alum.mit.edu> wrote:
> Cygnus has regularly delivered the code for major
> features to the EGCS tree, and I suspect quite a few
> major features appeared in the EGCS tree before they were
> shipped to Cygnus customers.

That would be an interesting suspicion to back up. I know
only of major features in the Cygnus tree that are not in
EGCS. Indeed there are features in GNUpro that have not
yet been put in EGCS. I am sure they all will be, just as
all internal GNAT developments are eventually made public.

Yes, of course Cygnus drops developments to EGCS, no one
is saying that they don't. It is just that, as in the case
of GNAT, there is usually a lag between the development
of new features and their dropping into EGCS.

Of course anyone could take the public version of GNAT and
do an EGCS like setup for it (the GNAT/Linux project has
some of these aspects). However, that is much more
effective if ACT not only drops new versions periodically,
but also develops fixes for the current public version
that can go there without waiting months for the next drop.
But that is what takes very substantial resources, and we
don't have the resources to do this yet, although we are
hoping to do it on a limited basis starting with 3.12 and
specifically for the GNAT/Linux project.

Right now, GNAT is more analogous to how Cygnus handles
GDB. That may change in the future, but Cygnus feels that
since it does almost all the work on GDB currently, it is
not obviously worth their while to invest resources in an
EGCS like effort for GDB. That may well change in the
future, if other parties start doing major work on GDB.

The question of the level of activity from other parties
is of course a critical one. Most of the agitation with
respect to GNAT has been from people who want to get early
versions for their own use, and unfortunately there has
been relatively little indepedent effort (in fact two
of the big independent contributors to the GNAT effort,
Doug Rupp for the DOS port, and Geert Bosch for the OS/2
port now work full time for ACT.

I must say that Markus Kuhn and others seem to be having
some success in getting the Linux/GNAT project going, and
an interesting first event a few weeks ago was the first
submission of a patch from this project that was adopted
into the GNAT sources since it was correct and helpful.
A small start, but an important one, and perhaps the
harbinger of a more active development and flow of bug
fixes and improvements in both directions. Realistically
one can expect more action in the area of bindings and
tools than in the compiler proper, but who knows?


 In fact, EGCS began
> as essentially a drop of the Cygnus internal development
tree.
>
> Of course, I am not a Cygnus employee so I can't really
speak for
> their proedures, but that is how it looks to me.
>
> Also, even though Cygnus provides the infrastructure for
EGCS and pays
> a number of employees to work on it, EGCS is not really
controlled by
> Cygnus, but rather by a steering committee on which
Cygnus employees
> constitute a minority.
>
>  - Maciej Stachowiak
>

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: EGCS & GNAT Was (Re: GNAT versions)

[Robert Dewar]

In article <37234DCA.A509D0D9@aasaa.ofe.org>,
  David Starner <dstarner98@aasaa.ofe.org> wrote:
> While it's on topic, since EGCS is now the official GCC
> tree, when will there be a version of GNAT that builds on
> EGCS? It would be much more convientent for those of use
> running EGCS.

3.12 technology is based on gcc 2.8.1 with a patch set,
as in previous releases.

Right now, a number of fixes and improvements are required
to EGCS to get it up to the level required by our
regression and QA testing (remember that, unlike the GCC
2.8.1 tree, changes have not been verified against our
test suite, and there are definitely some errors).

It is certainly our intention to migrate in the EGCS
direction, but it may take a while to complete this change
over, since clearly we cannot release an EGCS based version
that has significant regressions.

Just getting GNAT to run on EGCS is not that difficult.
Getting it to run perfectly, or at least as perfectly as
on GCC 2.8.1 will take some effort. How much we don't know
for sure yet.

If you are happy with a compiler that mostly works even if
it does not pass all our tests, then the patches for
running version 3.11p of GNAT under EGCS have already
been developed and posted by volunteers (in fact the first
set of such patches appeared about 24 hours after the
initial drop of 3.11p).

Robert Dewar
Ada Core Technologies

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: EGCS & GNAT Was (Re: GNAT versions)

[Robert Dewar]

In article <37234DCA.A509D0D9@aasaa.ofe.org>,
  David Starner <dstarner98@aasaa.ofe.org> wrote:

> I don't see problems arising from the EGCS
> snapshots being distributed; if you download a snapshot
> you know what you're getting.

Yes, but as our discussion has shown the EGCS snapshots
are not equivalent to our wavefronts. The GCC equivalent
of our wavefronts would be day by day drops of the internal
Cygnus development tree, something that Cygnus does not do
(for reasons we understand very well!)

As I said in an earlier message, if there was something
equivalent to EGCS for GNAT, that would be a quite
different situation but so far there is not, although
perhaps the GNAT/Linux project will develop in that
direction!


-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: EGCS & GNAT Was (Re: GNAT versions)

[Larry Kilgallen]

Reply-To: Kilgallen@eisner.decus.org.nospam
Organization: LJK Software
Lines: 16

In article <37234DCA.A509D0D9@aasaa.ofe.org>, David Starner <dstarner98@aasaa.ofe.org> writes:

> I still don't understand the great worries about GNAT beta versions
> getting distributed.

Neither do I, but people keep bringing this up in comp.lang.ada
as a political football, without anyone coming up with a single
example where there has been a problem with the current system
of releasing only software that has received the fuller testing.

As an Ada fan, I would much prefer the versions of Ada available
(from all sources) were as defect-free as possible, so those who
are not already firmly in the Ada camp will not get disuaded by
encountering defects.

Larry Kilgallen

Re: GNAT versions ( was :Ada compiler for PC?)

[Geoffrey KEATING]

Robert Dewar <robert_dewar@my-dejanews.com> writes:

> In article <t8zp3xy1w7.fsf@discus.anu.edu.au>,
>   Geoffrey KEATING <geoffk@discus.anu.edu.au> wrote:
> > This is certainly true for egcs, and IIRC even for gcc
> > there is anonymous CVS access.
> 
> But surely you do not think that the current EGCS snapshot
> does not represent the internal Cygnus development tree
> (actually you may think this, I quite often run into people
> under this remarkable illusion!

These people are no doubt confused; possibly they think that 'EGCS' is
the same thing as 'Cygnus GCC'.  These are usually the same people
that think that Cygnus runs the egcs project.

The current egcs snapshot doesn't represent _my_ internal development
tree, either, although I note that I'm down to only about four patches'
difference between my tree and current snapshots.  Similarly, Cygnus
is continually trying to reduce the difference between the egcs tree
and their version, if only because it makes it easier to update their
tree to the current egcs one.

-- 
Geoff Keating <Geoff.Keating@anu.edu.au>

Re: GNAT versions ( was :Ada compiler for PC?)

[bourguet]

In article <t8zp3xy1w7.fsf@discus.anu.edu.au>,
  Geoffrey KEATING <geoffk@discus.anu.edu.au> wrote:
> bourguet@my-dejanews.com writes:
[...]
> You mean, the wavefront and/or commercial releases of gnats are
> available by anonymous FTP, and the current tree can be obtained by
> anonymous CVS?  I'd be interested in more details, if so.

No. As far as I know wavefront version are not even available to customer
who don't have a need for (I.E. are blocked by a bug which does not have
a raisonable work around).

> This is certainly true for egcs, and IIRC even for gcc there is
> anonymous CVS access.
>
> [Before anyone asks, you can get egcs snapshots from egcs.cygnus.com,
> or any of its mirrors.  Releases are put in the same place.]

I've already been to egc.cygnus.com and saw announce like "cygnus donates
code to support processor X" "cygnus donates code to perform optimization
Y". Do you thinks these code are not in a development tree not publicly
accessible and have not been tested, perhaps by customers of cygnus?

I see no raisons to thinks this; more I see raisons for cygnus having an
internal development tree and giving sample of what is in this development
tree to customers.

-- Jean-Marc

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: EGCS & GNAT Was (Re: GNAT versions)

[Larry Kilgallen]

In article <7g0qj1$1td0$1@Mercury.mcs.net>, les@MCS.COM (Leslie Mikesell) writes:
> In article <1999Apr25.201259.1@eisner>,
> Larry Kilgallen <kilgallen@eisner.decus.org> wrote:

>>As an Ada fan, I would much prefer the versions of Ada available
>>(from all sources) were as defect-free as possible, so those who
>>are not already firmly in the Ada camp will not get disuaded by
>>encountering defects.
> 
> If the old versions are perfect, why is there ongoing development?

Of course some development is for new features.  A compiler that
complied perfectly with the RM could still be improved in optimization,
quality of error messages, etc.  There might be other defects going to
the absolute correctness of the compiler.  What there should never be
in a released version is DEFECTS THAT WERE PREVIOUSLY FIXED.

If you maintain that those internal "wavefront" releases cannot
introduce a regression, might it be the case that ACT wastes time
by running full regression tests before a release ?  If so, then
you can dismiss all the alleged "added value" Robert spoke about
from their extensive test battery and just make your own changes
to the compiler.

In fact, if you are so upset about the way ACT behaves, you can set
up a competing GNAT development effort.  When the time comes, however,
for me to pay money to a vendor for GNAT, and I have my choice between
a vendor who tests and a vendor who doesn't, I have no question about
which way my decision would go.

Larry Kilgallen

Re: EGCS & GNAT Was (Re: GNAT versions)

[Robert Dewar]

In article <7g0qj1$1td0$1@Mercury.mcs.net>,
  les@MCS.COM (Leslie Mikesell) wrote:
> It is mostly bickering about the meaning of the GPL.  As
> in how you can distribute even one bugfix copy of
> something with a license that ensures the right to source
> and the right to redistribute the source while witholding
> the source

There is nothing in the GPL that ever *requires* *anyone*
to distribute *anything* [other than the source for a
program for which they distribute the objects] under
*any* circumstances. Indeed a license which has any such
requirements is not regarded as a legitimate free software
license by the FSF (this is one of the contentions, the
open source software folk to do not make this distinction)

> and insisting that the recipient not redistribute?

Any such insistence would be a clear violation of the GPL.
For example, if you distributed GPL'ed software along with
a non-disclosure agreement which restricted the further
distribution, this would violate the GPL. You can
explicitly ask people not to do it, explaining why you
would prefer them not to redistribute, but that's only
a request. At ACT, we don't even make such explicit
requests.

I suspect there *are* situations that are murky, for
example surrounding the GCC work being done for Merced, but
at ACT we never have any restrictive requirements, since,
as noted above, these would clearly violate the GPL, and
you are right, to have such restrictions would make no
sense, which is why the GPL does not permit them.

Robert Dewar
Ada Core Technologies


  One or the other doesn't
> make sense.
>
> >As an Ada fan, I would much prefer the versions of Ada
available
> >(from all sources) were as defect-free as possible, so
those who
> >are not already firmly in the Ada camp will not get
disuaded by
> >encountering defects.
>
> If the old versions are perfect, why is there ongoing
development?
>
>   Les Mikesell
>    les@mcs.com
>

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: EGCS & GNAT Was (Re: GNAT versions)

[Robert Dewar]

In article <1999Apr26.074128.1@eisner>,
> In article <7g0qj1$1td0$1@Mercury.mcs.net>, les@MCS.COM
> (Leslie Mikesell) writes:

> > If the old versions are perfect, why is there ongoing
> >development?

First: no one said that the old versions are perfect, even
with respect to basic issues of conformance to the RM.
Every version of GNAT so far has had bugs. No amount of
testing can guarantee elimination of all bugs.

But more to the point, the great majority of our ongoing
development effort relates to new features, new tools,
new packages, improved performance etc.

Here is a brief excerpt from the features file for the
forthcoming 3.12 releases. I will post the whole of this
when we freeze the 3.12 sources, which has not happened
yet.

  The compiler is now built with options -gnatpn instead of
  -gnata. This means that the front end of the compiler is
  considerably faster, up to 2-3 times faster in some
  cases. The cases where you will see the biggest speed up
  are in -gnatc compilations with no code generation, or if
  very large specs are with'ed from smaller units.

  If pragma Suppress is used in the gnat.adc file, this now
  properly
  suppresses exceptions in all files compiled in the
  presence of this
  gnat.adc file (Suppress pragmas in gnat.adc were
  previously ignored,
  which is in accordance with the RM, but certainly not
  what is wanted!)

  On Digital Unix 4.0D, the run time now takes advantage of
  the full
  range of priorities (0 .. 63).

  In -gnatc mode, an existing up to date ali file is no
  longer destroyed.
  In particular this means that the -gnatc -gnatt
  compilations used by
  ASIS do not destroy existing ali files.

  A new switch -gnaty activates style checking features in
  the compiler.
  These roughly correspond to the checking done by the
  special internal
  -gnatg flag, except that -gnaty allows extensive choice
  of which checks
  are to be performed, and also allows parametrization,
  e.g. of the indent
  level that is enforced.

  The handling of aggregates has been optimized in many
  cases, generating
  more efficient code and less memory usage.

  The binder now generates an Ada package as the main
  program by default
  instead of a C program. The generated files are called
  b~xxx.ads/adb,
  where xxx is the name of the main program. The -C switch
  for both
  gnatbind and gnatlink can be used to get the old behavior
  of generating
  the main program in C.

This is just a small excerpt (the full features file
entry for 3.12 is about 200 lines long and growing), but
this should give you a feel for what continued development
is about.

Indeed our support activities these days are far less
focussed on fixing bugs than helping people use Ada 95
and GNAT features successfully -- though of course the
bug fixing activity is an important component still!

Robert Dewar
Ada Core Technologies

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GNAT versions ( was :Ada compiler for PC?)

[Per Bothner]

Maciej Stachowiak <mstachow@alum.mit.edu> wrote:
> ... I suspect quite a few major features appeared in the EGCS
> tree before they were shipped to Cygnus customers.

In article <7g0bdf$3q9$1@nnrp1.dejanews.com>,
Robert Dewar  <robert_dewar@my-dejanews.com> wrote:
>That would be an interesting suspicion to back up. I know
>only of major features in the Cygnus tree that are not in EGCS.

Well, I can't think of a single Gcc feature, major or otherwise,
that was in our standard customer release before being in EGCS.
Note:  I am talking about the standard GNUPro product;  not
contracted deliverables made to a specific customer.  (Obviously,
if somebody pays for a new port to an unannounced chip, we are
not going to put into Egcs before it is announced!)  But in
general, customers do *not* get major features before Egcs.

Since we merge *from* Egcs to our internal tree, rather than
vice versa, the check-in policy at Cygnus is:  Nothing gets
into our internal tree unless it is in Egcs *or* specially
marked as being Cygnus only or "sanitized".  That should make
it obvious that the default is to check things into Egcs
first or at the same time.

>Yes, of course Cygnus drops developments to EGCS, no one
>is saying that they don't. It is just that, as in the case
>of GNAT, there is usually a lag between the development
>of new features and their dropping into EGCS.

As I have pointed out, this is generally not true.
It may be worth pointing out that Dewar's company ACT is a
competitor of Cygnus, though we have mostly kept apart.

>Indeed there are features in GNUpro that have not
>yet been put in EGCS. I am sure they all will be, just as
>all internal GNAT developments are eventually made public.

Well, some parts of our product are non-free, and there are no
plans to change that (though the mix of what is free and what
is non-free will of course change).  Specifically, Source
Navigator (which is a separate product from GNUPro) is
non-free, as is gdbtk (which I believe *is* part of GNUPro).

>Right now, GNAT is more analogous to how Cygnus handles
>GDB. That may change in the future, but Cygnus feels that
>since it does almost all the work on GDB currently, it is
>not obviously worth their while to invest resources in an
>EGCS like effort for GDB. That may well change in the
>future, if other parties start doing major work on GDB.

I don't think it is proper for you to say what "Cygnus
feels".  At the very least put in a "my guess is ...".

The reason EGCS was started was because things has reached a
crisis point.  This happened while the official Gcc maintainer
was working for ACT.  Gdb has also had problems the last year
with openness and timely releases;  we are working on improving this.
A public cvs repository is one possibility (but I am not in the
gdb group, so I cannot speak for them).
-- 
	--Per Bothner
Cygnus Solutions     bothner@cygnus.com     http://www.cygnus.com/~bothner

Re: GNAT versions ( was :Ada compiler for PC?)

[Matthew Heaney]

Robert Dewar <robert_dewar@my-dejanews.com> writes:

> In article <7g3klg$26p$1@rtl.cygnus.com>,
>   bothner@cygnus.com (Per Bothner) wrote:
> > The reason EGCS was started was because things has
> > reached a crisis point.  This happened while the official
> > Gcc maintainer was working for ACT.
> 
> Since Cygnus likes to correct you when people say that
> Cygnus controls EGCS, let's clarify a little here. Richard
> Kenner in his role as the maintainer of GCC was working
> for New York University, not for ACT! That's an important
> distinction, just as important as the fact that EGCS is
> not controlled by Cygnus :-)


But was I supposed to infer from the comment that because Richard (the
GCC maintainer) was working for ACT (really, NYU), that that caused or
otherwise contributed to the crisis?

Re: GNAT versions ( was :Ada compiler for PC?)

[Stan Shebs]

bothner@cygnus.com (Per Bothner) writes:

> [...]  Gdb has also had problems the last year
> with openness and timely releases;  we are working on improving this.
> A public cvs repository is one possibility (but I am not in the
> gdb group, so I cannot speak for them).

The repository already exists, take a look at sourceware.cygnus.com.

							Stan Shebs
							Cygnus Solutions
							shebs@cygnus.com

Re: GNAT versions ( was :Ada compiler for PC?)

[Robert Dewar]

In article <7g3klg$26p$1@rtl.cygnus.com>,
  bothner@cygnus.com (Per Bothner) wrote:

> Well, I can't think of a single Gcc feature, major or
> otherwise, that was in our standard customer release
> before being in EGCS.

There have been some minor examples (at least you have
told customers this was the case), but in general, right,
the Cygnus policy is the same as the ACT one, anything in
a standard customer release is also made available publicly
at essentially the same time (in the case of ACT, there is
a small lag in time, simply because we prepare the customer
release first, if we had more resources, we could probably
eliminate even this couple of weeks lag).

> Note:  I am talking about the standard GNUPro product;
> not contracted deliverables made to a specific customer.

But actually this was the crux of the issue here in the
discussion with regard to GNAT. Not the issue of standard
releases, but the issue of special releases to specific
customers to meet specific contract requirements for these
customers.

> (Obviously, if somebody pays for a new port to an
> unannounced chip, we are not going to put into Egcs
> before it is announced!)

What I wonder about here is the secondary distribution.
Yes, as Cygnus you can give the product back to the chip
maker here, but presumably the chip maker cannot distribute
this to any third party with any kind of ND restrictions.
That seems an awkward situation.

For example, suppose that Cygnus did a Merced port for
Intel, then Intel could not give it to anyone else, e.g.
to develop a Linux for it, with non-disclosure restrictions
that would prevent them from distributing it.

One of the interesting points of the GPL is that it allows
ONE organization to do internal work in complete secrecy,
unhampered by GPL restrictions of any kind, but this
does not extend to two companies working together, because
every time they send a copy between them, it must be sent
in unrestricted manner.

I should say that I have discussed this point explicitly
with Richard Stallman, and he agrees with this assessment
of both the intent and effect of the GPL.


> I don't think it is proper for you to say what "Cygnus
> feels".  At the very least put in a "my guess is ...".

This was not a guess, it was based on a direct conversation
with Cygnus folks. Incidentally I did not intend it as a
criticism. If indeed most changes to GDB are done by Cygnus
it is perfectly reasonable for them to worry that opening
things up can cause extra work without sufficient return.
The reason EGCS works well on both sides is that although
it is true that most contributions to EGCS have been from
Cygnus, there are many important contributions from
elsewhere. So far this is not true of GDB, but we hope it
may be in the future, and Cygnus certainly shares this
(again I am not guessing, this is based on direct
interactions, but in any case it is an entirely reasonable
position, and is what I would have guessed in any case).
>

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GNAT versions ( was :Ada compiler for PC?)

[Robert Dewar]

In article <7g3klg$26p$1@rtl.cygnus.com>,
  bothner@cygnus.com (Per Bothner) wrote:
> The reason EGCS was started was because things has
> reached a crisis point.  This happened while the official
> Gcc maintainer was working for ACT.

Since Cygnus likes to correct you when people say that
Cygnus controls EGCS, let's clarify a little here. Richard
Kenner in his role as the maintainer of GCC was working
for New York University, not for ACT! That's an important
distinction, just as important as the fact that EGCS is
not controlled by Cygnus :-)


-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GNAT versions ( was :Ada compiler for PC?)

[Robert Dewar]

In article <m3u2u2t0xh.fsf@mheaney.ni.net>,
  Matthew Heaney <matthew_heaney@acm.org> wrote:

> But was I supposed to infer from the comment that because
> Richard (the GCC maintainer) was working for ACT (really,
> NYU), that that caused otherwise contributed to the
> crisis?

Well I can't speak for the intentions of the poster in
terms of what he intended you *infer*, but if you were
to draw such an inference it would be entirely incorrect.
There is quite a bit of history rewriting going on here,
but no one has said anything *quite* that outrageous, and
to be fair, I do not think that the original poster
intended you to draw such a conclusion, which would be
entirely false.

On the contrary, as with many other volunteer activities,
the critical thing is for the time of volunteers to be
supported adequately, and indeed ACT put substantial
resources into assisting with the continued maintenance
of GCC, something that benefited us, but also benefited
the larger GCC community.

A similar situation arises with Cygnus today, they invest
substantial resources to help with the maintenance of the
Cygnus tree, and those resources are enormously helpful to
the gcc community at large, not just to Cygnus.

That leads some people to harbour suspicions
that Cygnus really secretly controls the process and
somehow twists it to serve their own ends, but really this
is quite absurd, and does not correspond at all to the
reality of the situation.

To draw the inference you suggest here is equally absurd
(as well as being quite unfair to those involved!)

Robert Dewar
Ada Core Technologies

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GNAT versions ( was :Ada compiler for PC?)

[Robert Dewar]

In article
<m2so9lccbw.fsf@mailhost.neuroinformatik.ruhr-uni-bochum.de

> It *does* correspond to the reality of the situation as
> Cygnus invests considerable manpower in the process.  In
> a way, they are preparing well-trodden paths that are
> convenient for their purposes by investing appropriate
> resources.  If they are not interested in some direction
> of work, they'll leave it to other parties.

Yes, of course, anyone putting work into gcc is inclined
to put in things that they find interesting and useful,
and who ever puts in more effort will have more influence.
But there is nothing improper, suspicious, or in any way
negative about that.

A similar situation occurs with GNAT. Since ACT makes
almost 100% of the changes to GNAT currently (we only
very rarely get patches submitted, almost all of which
have been immediately incorporated), ACT gets to make
the decisions on which features go in etc. This is of
course based primarily on the needs of our paying
customers.

(two of the the most notable volunteer contributions to
GNAT were the DOS port by Doug Rupp, and the OS/2 work
by Geert Bosch, and not only did we acquire this work,
but Doug and Geert now work full time for ACT/ACTE.)

If the GNAT/Linux project ends up encouraging more useful
contributions, particularly in the tools and binding area,
then it will begin to have a significant influence on the
direction of the GNAT development, and that is welcome all
round as far as I am concerned.

There are some other interesting developments such as
David Botton's work on COM interfacing on NT, and ACT
is definitely interested in working closely with anyone
doing useful development that is GNAT related.

Robert Dewar
Ada Core Technologies



-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GNAT versions ( was :Ada compiler for PC?)

[David Kastrup]

Robert Dewar <robert_dewar@my-dejanews.com> writes:

> A similar situation arises with Cygnus today, they invest
> substantial resources to help with the maintenance of the
> Cygnus tree, and those resources are enormously helpful to
> the gcc community at large, not just to Cygnus.
> 
> That leads some people to harbour suspicions
> that Cygnus really secretly controls the process and
> somehow twists it to serve their own ends, but really this
> is quite absurd, and does not correspond at all to the
> reality of the situation.

It *does* correspond to the reality of the situation as Cygnus invests
considerable manpower in the process.  In a way, they are preparing
well-trodden paths that are convenient for their purposes by investing
appropriate resources.  If they are not interested in some direction
of work, they'll leave it to other parties.

While I find no wrong with that, it would be foolish to deny that the
willingness to invest a large amount of work in GCC *does* have a
large influence on what direction the compiler development is taking.

If a few larger vendors realized this somewhat more (Intel seems to
have waken up to it somewhat recently), this could only help GCC
development.  CPU manufacturers actually should have an even larger
interest as Cygnus in getting actively involved in GCC.  It lends
large value improvements to their products with minimal costs.


-- 
David Kastrup                                     Phone: +49-234-700-5570
Email: dak@neuroinformatik.ruhr-uni-bochum.de       Fax: +49-234-709-4209
Institut f�r Neuroinformatik, Universit�tsstr. 150, 44780 Bochum, Germany

Re: GNAT versions ( was :Ada compiler for PC?)

[Per Bothner]

I don't want to say too much here, because I don't remember or know
all the issues [in starting Egcs].  But it is no secret that Kenner
was not working out as Gcc maintainer.  This was not due to any lack
of technical skills.  I think the biggest problems were a lack a
management skills, and an unwillingness to delegate.  And as Dewar
says:  being Gcc maintainer is a tremendous time sink, and doing a
good job while maintaining a "day job" is almost impossible.

I don't think it was a major factor that Kenner was also working for
ACT with the associated possible conflict of interest.  However,
there was at least the perception that in a few case he made some
questionable decisions based on concerns about GNAT.  For example,
C++ exception handling was delayed because Kenner had his own ideas
as to how it "ought" to be done.  (I don't know what the technical
issues were.  Maybe he was just too much of a perfectionist - this
is certainly a problem I'm all too familar with in myself.)
-- 
	--Per Bothner
Cygnus Solutions     bothner@cygnus.com     http://www.cygnus.com/~bothner

Re: GNAT versions ( was :Ada compiler for PC?)

[Robert Dewar]

In article <7gbjhg$s98$1@rtl.cygnus.com>,
  bothner@cygnus.com (Per Bothner) wrote:

> I don't think it was a major factor that Kenner was also working for
> ACT with the associated possible conflict of interest.  However,
> there was at least the perception that in a few case he made some
> questionable decisions based on concerns about GNAT.  For example,
> C++ exception handling was delayed because Kenner had his own ideas
> as to how it "ought" to be done.  (I don't know what the technical
> issues were.  Maybe he was just too much of a perfectionist - this
> is certainly a problem I'm all too familar with in myself.)

It is probably fair to say that Kenner is a perfectionist. He thinks
that things should be documented (Per says he does not know the issues,
perhaps he should check :-) But one of the major issues with exception
handling was indeed documentation, and in fact a special deal was done
to put this into gcc without the documentation, accepting instead a
promise that it would be done later!

The concerns about exceptions had nothing at all to do with GNAT. This
is simply a guess on Per's part, who as he says, does not know the
issues. The concern was on technical issues (such as thread safety,
which of course are language independent), and on the documentation
issues.

I think Per's continued implications that perhaps Richard was working
deliberately against Cygnus interests because ACT was a competitor are
pretty bogus on the face of it. ACT's interests have been very
specifically in Ada (consider the name of the company!) And as far as
we know Cygnus has had zero interest in Ada.

In practice, the issue was indeed to a considerable extent one of
too much work for any one person. ACT funded a lot of Richard Kenner's
time to work on gcc in practice, so that it could indeed be his day
job, but even so, there is a lot for one person to do, and furthermore,
it is pretty hard to please everyone in a situation like this.

But in general I would advise Per to investigate thoroughly the
circumstances of both EGCS at the start, and the exception issues,
before making guesses. Even with the disclaimers that he does not
know the issues, such guesses can be misleading.

Robert Dewar
Ada Core Technologies

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GNAT versions ( was :Ada compiler for PC?)

[Mike Stump]

In article <7gbjhg$s98$1@rtl.cygnus.com>,
Per Bothner <bothner@cygnus.com> wrote:

>However, there was at least the perception that in a few case he made
>some questionable decisions based on concerns about GNAT.  For
>example, C++ exception handling was delayed because Kenner had his
>own ideas as to how it "ought" to be done.

Here is my take...

I don't think any of this was related to GNAT.  My take is just a
simple conflict of styles.  It only delayed EH by 2-7 years.  My
choice, would have been to not delay it, and just get it working, then
working well, then well ported, then clean up the interfaces and
documentation.

I resisted doing it his way, and he resisted accepting it my way.

I don't yet understand the benefit of delaying EH, maybe in time I
might...

In retrospect, I think it might have been easier to do the entire
thing, port it to all the major platforms, work out the details, clean
it up, document it, then submit it after the fact.  What I wanted to
do, was expose the source code control system to my learning, so as to
preserve the record of what was learned.  One can do up the whole
thing first, then check it in en-mass, but what I don't like about
that, is there is no history of why.  It _feels_ less open to me.  I
think there are more benefits by having it more _open_.

Re: EGCS & GNAT Was (Re: GNAT versions)

[Mike Stump]

In article <7g2798$ntf$1@nnrp1.dejanews.com>,
Robert Dewar  <dewar@gnat.com> wrote:
>In article <1999Apr26.074128.1@eisner>,
>But more to the point, the great majority of our ongoing
>development effort relates to new features, new tools,
>new packages, improved performance etc.

I have a question, will ACT be able to help with the egcs community,
and help us integrate in teh Ada frontend into egcs, or do we have to
go it alone?

I personally would like to see Ada and Pascal integrated, tested (to
some extent), and then egcs released as gcc 3.0.

Re: EGCS & GNAT Was (Re: GNAT versions)

[Robert Dewar]

In article <FB8F5o.DIH@kithrup.com>,
  mrs@kithrup.com (Mike Stump) wrote:

> I have a question, will ACT be able to help with the egcs community,
> and help us integrate in teh Ada frontend into egcs, or do we have to
> go it alone?
>
> I personally would like to see Ada and Pascal integrated, tested (to
> some extent), and then egcs released as gcc 3.0.

Our priority is to get 3.12 out of the door first, 3.12 will still
be gcc 2.8.1 based. FOllowing that we will start serious work on
the EGCS integration. It is really something that ACT has to be
involved with, because access to our regression suite is crucial
for locating and dealing with EGCS regressions.

We are setting up to do the full regression run now, and should have
results soon. Certainly the 3.12 source base is the right one to
include in the next EGCS release.

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GNAT versions ( was :Ada compiler for PC?)

[Robert Dewar]

In article <FB8DLq.BH6@kithrup.com>,
  mrs@kithrup.com (Mike Stump) wrote:
> In article <7gbjhg$s98$1@rtl.cygnus.com>,

> I don't think any of this was related to GNAT.  My take is just a
> simple conflict of styles.  It only delayed EH by 2-7 years.  My
> choice, would have been to not delay it, and just get it working, then
> working well, then well ported, then clean up the interfaces and
> documentation.


That certainly seems backwards to me, for me the sequence of
software production goes like:

design and document the interfaces
refine the interfaces till they are correct
then implement

I find the style of doing things backwards worrisome, though I
realize it is common enough, especially in the C community.

We certainly follow this procedure internally at ACT. If someone
thinks a new feature is useful, then first we have a general
discussion of the idea, then we design the feature (e.g. package
specs are produced and reviewed), then we do the implementation
of the bodies. I doubt this description of normal software
design procedures seems strange, at least I would expect it
was entirely familiar to the comp.lang.ada users reading this!

The trouble with doing things backwards is that you are realying
FAR too heavily on testing alone as the criterion for correctness.
Even if there were a comprehensive and systematic test suite,
including significant amounts of code, testing is never sufficient
on its own to ensure that your design is correct.

I must say I am a little puzzled by Mike's reluctance to document
things. If EH was really delayed 7 years because of this (I think
that's a bit of an exaggeration :-) it is a pity ...

As to whether one should bend the rules and allow things to be
implemented without documentation, and instead hope for documentation
to appear later, it's hard to say. Sometimes you DO have to make
this compromise, but it often does not work out well. The trouble
is that many people who have this code-now, document-later approach
have a lot of trouble coming through with the documentation later.

For me personally, writing well documented code is a pleasure, sort
of like writing a text book that students like to read. If code is
accessible and easy to understand, and good documentation is part
of the reason, then that is a pleasing achievment in itself. I get
the impression that a lot of programmers don't share this view, and
find documentation a nuisance -- too bad.

P.S. I still find the documentation of the exception handling stuff
in GCC inadequate. This may be just lack of skill on my part in
reading C code of this kind (clearly there is an expectation that
you figure out some of the details of the specification by looking
at the details of the coding, which I find uncomfortable.

One unfortunate consequence is that the GNAT exception handling has
been developed entirely independently, and I suspect that there
could be more merging, though I am not sure of this, because to have
an exception handling mechanism usable by both C++ and Ada 95, a
perfectly reasonable goal, would require careful examination of
the specifications *before* starting to implement.

Anyway, I do agree with Mike on one important point, the exception
handling requirements for GNAT had nothing at all to do with the
development of EH handling in GCC!

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GNAT versions ( was :Ada compiler for PC?)

[Mike Stump]

>I still find the documentation of the exception handling stuff
>in GCC inadequate.

Well, feel free to donate improvements to the documentation, gcc
always welcomes clarifications and more complete explanations.  :-)
I'd be happy to answer your questions about it.

>One unfortunate consequence is that the GNAT exception handling has
>been developed entirely independently,

Well, I take exception to the word entirely.  There was a mailing list
called eh@cygnus.com, and we discussed tons of details and issues,
from many language perspectives, including Ada and C++.

This experience shaped and drove my implementation, the implementation
that now exists in gcc.

I would anticipate that it did the same for the Ada implementation.
If true, they are closely related.

>I suspect that there could be more merging,

Sure, each front end can replicate it's own exception scheme[1], but
this is a poor design.  A better design is where the common aspects of
the design are shared by the frontends as facilities in the backend.
C++ and Java already share EH.  Chill needs to be converted, but there
is no pressing need, as not to many people use or maintain it.

:-)  Yes, I know you already knew that.  I say it, just to nag a little,
in hopes that Ada will embrace the functionality that is now there and
extend it.

>because to have an exception handling mechanism usable by both C++
>and Ada 95, a perfectly reasonable goal, would require careful
>examination of the specifications *before* starting to implement.

:-)  Yes, and that was done to a large extent.  The missing pieces are
meant to slot into the existing design, in a natural way.


1 - Actually, my experience is that this is not possible, that is
    _why_ EH is now in gcc.  The frontend implementation could only
    go so far, then things like flow, really wanted to know more and
    understand more.

Re: GNAT versions ( was :Ada compiler for PC?)

[Robert Dewar]

In article <FBAEH2.9C0@kithrup.com>,
  mrs@kithrup.com (Mike Stump) wrote:

> Well, I take exception to the word entirely.  There was a mailing list
> called eh@cygnus.com, and we discussed tons of details and issues,
> from many language perspectives, including Ada and C++.

This was in my response to the claim that unfortunately the exception
handling in Ada had been developed entirely independently of the
mechanism in C++.

I did the design and implementation of the exception handling in
Ada, so I am pretty familiar with what was and was not done. Yes
there were early discussions about trying to deal with commonality
between the languages, but unfortunately these did not result in
a common facility. I tried to understand what had been done for
C++ but failed. Others here at ACT are still trying to do more
merging here, but it is not easy. I found no high level interface
oriented description of the mechanism used for C++. Perhaps I
simply did not look hard enough, or just did not know how to read
the code correctly. I find the back end of GCC rather difficult
to navigate (and indeed in a recent email Per Bothner (hope I
remembered the spelling right) claimed that the entire gcc compiler
was ill-documented so why single out the exception handling). I
find that position FAR too pessimistic, and indeed in general the
gcc backend is documented a lot better than many proprietary
compilers with which I am familiar, but in this particular case,
we have not yet achieved the ideal of merging the exception handling
of C++ and Ada. I believe this could only have been achieved if we
had started with a fully documented high level design.

Again, this may simply be a different way of working, here at ACT,
we really aren't very good at roaming around code and figuring out
what is going on, we depend on a very structured design approach, so
there is undoubtedly a bit of a culture clash :-) :-)


-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GNAT versions ( was :Ada compiler for PC?)

[Andi Kleen]

Robert Dewar <robert_dewar@my-dejanews.com> writes:

> I did the design and implementation of the exception handling in
> Ada, so I am pretty familiar with what was and was not done. Yes
> there were early discussions about trying to deal with commonality
> between the languages, but unfortunately these did not result in
> a common facility. I tried to understand what had been done for
> C++ but failed. Others here at ACT are still trying to do more
> merging here, but it is not easy. I found no high level interface
> oriented description of the mechanism used for C++. Perhaps I
> simply did not look hard enough, or just did not know how to read
> the code correctly. I find the back end of GCC rather difficult
> to navigate (and indeed in a recent email Per Bothner (hope I
> remembered the spelling right) claimed that the entire gcc compiler
> was ill-documented so why single out the exception handling). I
> find that position FAR too pessimistic, and indeed in general the
> gcc backend is documented a lot better than many proprietary
> compilers with which I am familiar, but in this particular case,
> we have not yet achieved the ideal of merging the exception handling
> of C++ and Ada. I believe this could only have been achieved if we
> had started with a fully documented high level design.

As a first impression the GNAT 3.11 "zero-cost" EH implementation 
seems to usually generate smaller unwind tables than the egcs C++ 
EH implementation

(IMNSHO that is the main problem with the current egcs EH - too much
executable bloat)

It would be great if the best points of both could be merged.

-Andi

Re: GNAT versions ( was :Ada compiler for PC?)

[Robert Dewar]

In article <m3so9aci9g.fsf@fred.muc.de>,
  Andi Kleen <ak@muc.de> wrote:

> As a first impression the GNAT 3.11 "zero-cost" EH implementation
> seems to usually generate smaller unwind tables than the egcs C++
> EH implementation
>
> (IMNSHO that is the main problem with the current egcs EH - too much
> executable bloat)
>
> It would be great if the best points of both could be merged.
>
> -Andi


I suspect this comparing apples and oranges, I am not sure what target
you are on, but I think you may be comparing the stack unwinding tables
of gcc with the static exception tables of GNAT, but maybe not.
Certainly the approach used for finalization in GNAT will use less
table space than the approach used in G++ but who knows if that is
significant.

As far as unwinding goes, on DEC Unix, we use the standard DEC Unix
library calls for unwinding, and on SGI, we use the standard SGI
routines. On the x86, we do prolog interpretation, which is very
messy but is highly space efficient. On Solaris, we plan to use the
standard gcc mechanism, but have not figured it out yet.

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GNAT versions ( was :Ada compiler for PC?)

[Mike Stump]

In article <7gs0d5$lvh$1@nnrp1.deja.com>,
Robert Dewar  <robert_dewar@my-dejanews.com> wrote:

>I tried to understand what had been done for C++ but failed. Others
>here at ACT are still trying to do more merging here, but it is not
>easy.

I'd be happy to answer any questions.

>I found no high level interface oriented description of the
>mechanism used for C++.

>I believe this could only have been achieved if we had started with a
>fully documented high level design.

You make it sound like it can't be achieved.  :-(

Now, back to the first part...  There are roughly three routines to
call, one starts a region, one ends it, and the last to emit the
region handlers.  My hope was it would be easy to understand/use.

/* Start an exception handling region.  All instructions emitted after
   this point are considered to be part of the region until
   expand_eh_region_end () is invoked.  */

extern void expand_eh_region_start		PROTO((void));

/* End an exception handling region.  The information about the region
   is found on the top of ehstack.

   HANDLER is either the cleanup for the exception region, or if we're
   marking the end of a try block, HANDLER is integer_zero_node.

   HANDLER will be transformed to rtl when expand_leftover_cleanups ()
   is invoked.  */

extern void expand_eh_region_end		PROTO((tree));

/* Called from expand_exception_blocks and expand_end_catch_block to
   expand and pending handlers.  */

extern void expand_leftover_cleanups		PROTO((void));

The first two are meant to be fairly easy to use, maybe you can
explain what you found hard to use about them, and I can clarify.

The last is somewhat magical, and really should be buried deeper into
the backend.  Maybe if we can move the exception specifications into
the backend, the reason for not having it in the backend can be
removed.  But, if one reads what java does:

void
emit_handlers ()
{
  if (catch_clauses)
    {
      rtx funcend = gen_label_rtx ();
      emit_jump (funcend);

      emit_insns (catch_clauses);
      expand_leftover_cleanups ();

      emit_label (funcend);
    }
}

(thanks Per), you can just plug this in (or move it into the backend,
fix java to not include it), and just put in a call to emit_handlers.

That is the $0.05 tour.  :-) Now, admittedly, there are a few more
concepts to understand, like the cleanups in
TARGET_EXPR/expand_decl_cleanup _are_ run on an exception, for
example, but not that many more.

Re: GNAT versions ( was :Ada compiler for PC?)

[Robert Dewar]

In article <FBCAHD.Fqw@kithrup.com>,
  mrs@kithrup.com (Mike Stump) wrote:
> In article <7gs0d5$lvh$1@nnrp1.deja.com>,
> Robert Dewar  <robert_dewar@my-dejanews.com> wrote:
>
> >I tried to understand what had been done for C++ but failed. Others
> >here at ACT are still trying to do more merging here, but it is not
> >easy.
>
> I'd be happy to answer any questions.

What I would really look for here is a comprehensive document, or a pointer
to documentation within the source, that would answer all the questions.
The trouble is that it is quite clear that the current interface will not
meet all the Ada requirements. The differences are partly at the detail
level, partly more fundamental. But we need to be discussing changes to
an existing interface document under some kind of configuration control.
At least that's the way we work :-)

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GNAT versions ( was :Ada compiler for PC?)

[Pascal F. Martin]

In article <7gpsrd$qc7$1@nnrp1.dejanews.com>,
	Robert Dewar <robert_dewar@my-dejanews.com> writes:
> In article <FB8DLq.BH6@kithrup.com>,
>   mrs@kithrup.com (Mike Stump) wrote:
> 
> That certainly seems backwards to me, for me the sequence of
> software production goes like:
> 
> design and document the interfaces
> refine the interfaces till they are correct
> then implement
> 

This is an idealistic view, as proving an interface correct is not (yet)
exact science. The mechanical and engineering science use a different
sequence which I like very much, as it copes with real life problems:

design and document the interfaces
refine the interfaces till they are accepted by peers
implement (this may infere a few changes in the interfaces)
document the interfaces "as built".

-- 

Pascal F. Martin.

Re: GNAT versions ( was :Ada compiler for PC?)

[Robert Dewar]

In article <J3YY2.2192$2j3.3654@clnws01.we.mediaone.net>,
  pmartin@mail.earthlink.net (Pascal F. Martin) wrote:
> This is an idealistic view, as proving an interface correct is not
(yet)
> exact science. The mechanical and engineering science use a different
> sequence which I like very much, as it copes with real life problems:
>
> design and document the interfaces
> refine the interfaces till they are accepted by peers
> implement (this may infere a few changes in the interfaces)
> document the interfaces "as built".

Sure some iteration is required, but it can often be minimized,
and you can come very close to the ideal.

For example, for one project I did for Honeywell, a full real time
executive, with thread support, synchronization primitives, integral
debugger etc, was done by FIRST writing the user reference manual
documenting the API in completely detail, THEN writing the executive
(some 50,000 lines of assembly code). very few changes were required
to the manual, almost all stylistic copy-editing stuff (they hired
an English PhD to look over the manual and I had to let her make
a *few* changes to avoid her completely destroying important technical
stuff :-)

But in any case our discussion here is over detail. Either approach
is very far from the "implement first, document alter if at all"
method :-)

Robert Dewar


-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GNAT versions ( was :Ada compiler for PC?)

[bill]

In article <7h2ig6$apr$1@nnrp1.deja.com>, Robert says...
 
>
>Sure some iteration is required, but it can often be minimized,
>and you can come very close to the ideal.
>

>For example, for one project I did for Honeywell, a full real time
>executive, with thread support, synchronization primitives, integral
>debugger etc, was done by FIRST writing the user reference manual
>documenting the API in completely detail, THEN writing the executive
>(some 50,000 lines of assembly code). very few changes were required
>to the manual,
 
I also like the document first, write code second.

But many time, after development starts, the API needs to be 
changed, becuase others wants it changed. it is not in your 
control. If you design your API one way, and no one complains then, 
but one week later, manager asks you to add something to the API or
change it becuase they need some new functionality from your 
system, what are you to do?

I found that when this happens, it is almost always becuase people did not
spend too much time on specifications and on design stage. This happens
also in places where little communication between engineers is the common
culture in the work place. And in places where software managers have no
clue about software, yet they are in charge of large software development
projects. 

I've had a software manager who had no idea what a source control system is,
and this was in a major US company. His idea of source control is to copy
your files to separate hard drive.
  
I've worked in places, where design changes were nade based on quick 
talk over the coffe station at work, or some remark made between 
2 programmers walking across each other in the hallway.

I've worked in places, where programmers do not talk to each others working
on the same project. I've even worked on projects where programmers do not
talk period.  It is not surprising that changes in API are always occuring
in these places. These places are some of the major companies in the US,
well known names.

It seems the most common way of writing code nowadays is this:

  -- Below is a simplified version of how software is written
  -- nowadays in some of the major companies in the US. 


   To make every one think they are doing things right,
   quickly write some design thing. email it. then goto write_code.

   (of course no one will read and comment on the design send,
    since everyone else is busy debugging, but it is done so one can say
    they have done 'design').


write_code:
   write more code.
   see if that does what you think it needs to do
   if not then
      goto write_code
   else
      done: email the program to customer.
      cutomer emails back a bug complaint.
      goto write code.
    end if


Bill

Re: GNAT versions ( was :Ada compiler for PC?)

[Richard Kenner]

In article <7gbjhg$s98$1@rtl.cygnus.com> bothner@cygnus.com (Per Bothner) writes:
>However, there was at least the perception that in a few case he made some
>questionable decisions based on concerns about GNAT.  For example,
>C++ exception handling was delayed because Kenner had his own ideas
>as to how it "ought" to be done. 

Well, since Mike Stump, the author of the delayed code in question,
didn't see it as an issue of concerns with GNAT either then or now,
I'm not sure what you mean by "perception".

I want to add to some of the comments that Robert Dewar made later in this
thread and to complete the history here.

A long time ago (perhaps 1992), there was a set of meetings held at
NYU on the topic of language-independent exception handling design.
At those meetings were folks representing Ada, C++, and Modula.  We
spent a long time trying to solve the problem of how to implement a
language- and machine-independent exception handling model, which was
and is a very hard technical problem.

These discussions continued on an EH list for a while after that.
Subsequently, Robert wrote up a document that contained a draft
specification for such a system, which was meant as a basis for
implementation.

Clearly, once implementation started, we would expect this specification
to have to be updated as people learned more about the process.  But
nobody seemed to have time to do the implementation.

Then an exception handling mechanism for GCC "suddenly" appeared
*without* there being any revisions or details added to this document.

One of the critical things I needed to know was whether or not the
basic mechanism embodied in this implementation was or was not based
on this document and, if it was, what modifications were made to those
details in the process.  This is a documentation issue and, as Mike
correctly points out, was the main source of the dispute between he
and I.

I wanted to have the code documented, not just at the detailed level,
but to have a specification of the basic strategy being used to
implement machine- and language-independent exception processing.
After a lot of back and forth (lasting at least a year, probably
more), RMS and I agreed to install the code and received in return a
promise for such documentation to be written.  This was a number of
years ago and, so far as I know, there has never been an update of the
Robert's design document with the details of the particular
implementation.

With hindsight, clearly what happened should have been avoided.  But,
like Mike, I'm not sure *how* it could have been avoided.  Perhaps we
could have done something similar to what EGCS was originally and set
up a parallel fork for the exception handling code where people could
work on improving and finishing it, but it would stay out of the GCC
releases until it was complete and documented.  The problem with
approach, of course, is that people would want to *use* the code (in
both C++ and Ada) and those requirements would essentially cause this
code fork to be its own "release" in much the same manner as EGCS was
in its early days and divert the already-limited set of testers to
testing *two* releases.

How to deal with situations like this is, I believe, one of the more
critical issues in open software development and I have no solutions
to offer: history tells us what *doesn't* work, but I'm unaware of any
solutions that do.

Re: GNAT versions ( was :Ada compiler for PC?)

[Nick Roberts]

Richard Kenner wrote ...
[...]
|A long time ago (perhaps 1992),

I keep getting an image of Richard's post scrolling away from me in 3D,
against a starfield background, accompanied by a rousing march.

"A long time ago, in a galaxy far away ... there was a trade dispute ..."


:-)
Nick

== Stay Cool, Hang Tough, Bath Often ==