What is the relationship between GCC and Cygnus / Red Hat?

What is the relationship between GCC and Cygnus / Red Hat?

[Gerald Pfeifer]

I believe by having the following entry in our FAQ, we actually reach the
opposite of what it was originally ment to -- namely calm fears of Red Hat
controlling GCC -- and I believe the situation has sufficiently changed
since then, so I would like to remove it.

However, this is my personal opionion, so I'd like to hear what others
think about it, especially folks from Red Hat and competing companies!

Gerald

Index: faq.html
===================================================================
RCS file: /cvs/gcc/wwwdocs/htdocs/faq.html,v
retrieving revision 1.191
diff -u -3 -p -r1.191 faq.html
--- faq.html	9 May 2003 14:29:12 -0000	1.191
+++ faq.html	9 May 2003 14:30:06 -0000
@@ -30,7 +30,6 @@ Information page</a>.</p>
   <li><a href="#general">General information</a>
   <ol>
      <li><a href="#gcc">What is the relationship between GCC and EGCS?</a></li>
-     <li><a href="#cygnus">What is the relationship between GCC and Cygnus / Red Hat?</a></li>
      <li><a href="#open-development">What is an open development model?</a></li>
      <li><a href="#support">How do I get a bug fixed or a feature added?</a></li>
      <li><a href="#platforms">Does GCC work on my platform?</a></li>
@@ -94,25 +93,6 @@ official GCC maintainers. The net result
 carries forward GCC development under the ultimate control of the
 <a href="steering.html">GCC Steering Committee</a>.</p>

-
-<hr />
-<h2><a name="cygnus">What is the relationship between GCC and Cygnus / Red Hat?</a></h2>
-
-<p>It is a common mis-conception that Red Hat controls GCC either
-directly or indirectly.</p>
-
-<p>While Red Hat does donate hardware, network connections, code and
-developer time to GCC development, Red Hat does not control GCC.</p>
-
-<p>Overall control of GCC is in the hands of the
-<a href="steering.html">GCC Steering Committee</a>
-which includes people from a variety of different organizations and
-backgrounds.  The purpose of the steering committee is to make
-decisions in the best interest of GCC and to help ensure that no
-individual or company has control over the project.</p>
-
-<p>To summarize, Red Hat contributes to the GCC project, but does not exert
-a controlling influence over GCC.</p>

 <hr />
 <h2><a name="open-development">What is an open development model?</a></h2>

Re: What is the relationship between GCC and Cygnus / Red Hat?

[Gabriel Dos Reis]

Gerald Pfeifer <pfeifer@dbai.tuwien.ac.at> writes:

| I believe by having the following entry in our FAQ, we actually reach the
| opposite of what it was originally ment to -- namely calm fears of Red Hat
| controlling GCC -- and I believe the situation has sufficiently changed
| since then, so I would like to remove it.
| 
| However, this is my personal opionion, so I'd like to hear what others
| think about it, especially folks from Red Hat and competing companies!

I do not have any strong opinion on the issue, however I would like to
point out that removing that entry from the FAQ may be interpreted as
removing an entry that is no longer true, i.e. there would be a take
over from Red Hat -- I know that is not true, but the mere act of
removing it may be interpreted that way.  

Is there a way (other than removal) to clarify the situation?

-- Gaby

Re: What is the relationship between GCC and Cygnus / Red Hat?

[Michael Matz]

Hi,

On Fri, 9 May 2003, Gerald Pfeifer wrote:

> However, this is my personal opionion, so I'd like to hear what others
> think about it, especially folks from Red Hat and competing companies!

I think also it should be removed.  Certainly mentioning that Red Hat
provides resources in that paragraph is bogus, as others do that too.
I.e. it singles out Red Hat on the positive and negative side, and I too
can't see the value in that paragraph anymore (if I ever had).


Ciao,
Michael.

Re: What is the relationship between GCC and Cygnus / Red Hat?

[Russ Allbery]

Gerald Pfeifer <pfeifer@dbai.tuwien.ac.at> writes:

> I believe by having the following entry in our FAQ, we actually reach
> the opposite of what it was originally ment to -- namely calm fears of
> Red Hat controlling GCC -- and I believe the situation has sufficiently
> changed since then, so I would like to remove it.

> However, this is my personal opionion, so I'd like to hear what others
> think about it, especially folks from Red Hat and competing companies!

I certainly have no objections to removing it entirely; GCC and in general
GNU toolchain development hasn't looked even remotely dominated by Cygnus
or Red Hat to me for a long, long time now.  But if that isn't acceptable
for whatever reason, an alternate suggestion:

> -<p>It is a common mis-conception that Red Hat controls GCC either
> -directly or indirectly.</p>
> -
> -<p>While Red Hat does donate hardware, network connections, code and
> -developer time to GCC development, Red Hat does not control GCC.</p>

Remove these two paragraphs.  I think they come across as alarmist
(unnecessarily so), since this is no longer a common conception (at least
as far as I'm concerned).

> -<p>Overall control of GCC is in the hands of the
> -<a href="steering.html">GCC Steering Committee</a>
> -which includes people from a variety of different organizations and
> -backgrounds.  The purpose of the steering committee is to make
> -decisions in the best interest of GCC and to help ensure that no
> -individual or company has control over the project.</p>

Leave this.

Add:

  <p>Red Hat does donate hardware, network connections, code and
  developer time to GCC development, as do many other companies.  Such
  contributions are coordinated by the GCC Steering Committee.</p>

or something similar to continue to give credit where credit is due and to
head off any feelings that Red Hat's involvement is being masked somehow.

Then, this too can be dropped:

> -<p>To summarize, Red Hat contributes to the GCC project, but does not exert
> -a controlling influence over GCC.</p>

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

Re: What is the relationship between GCC and Cygnus / Red Hat?

[Joe Buck]

On Fri, May 09, 2003 at 05:05:07PM +0200, Michael Matz wrote:
> I think also it should be removed.  Certainly mentioning that Red Hat
> provides resources in that paragraph is bogus, as others do that too.
> I.e. it singles out Red Hat on the positive and negative side, and I too
> can't see the value in that paragraph anymore (if I ever had).

Agreed; the purpose was historical (it was widely alleged in the egcs days
that egcs was a plot by Cygnus to take over gcc).

Best action is to just drop the question, since it is no longer frequently
asked.

Re: What is the relationship between GCC and Cygnus / Red Hat?

[DJ Delorie]

I think we should keep it, but replace "Cygnus/Red Hat" with
"$COMPANY" as it applies equally to many companies.

Re: What is the relationship between GCC and Cygnus / Red Hat?

[Tom Lord]

	jbuck:

	> Agreed; the purpose was historical (it was widely alleged in
	> the egcs days that egcs was a plot by Cygnus to take over
	> gcc).

Isn't there a more accurate balance between the conspiracy theory take
and the conspiracy-denial take?

I think you are correct that Cygnus did not plot to "take over" gcc.

I also think that many core contributors to gcc worked, via egcs, to
end the then-traditional FSF-style of control over GCC.

There _was_ a power struggle there.  It'd be wrong to accuse Cygnus of
plotting.  It'd be wrong to paint it as an underhanded plot (it seemed
to bend over backwards to be just the opposite, imo).  It'd be wrong
to dredge up any old interpersonal issues among the personalities
involved.

But it was a power struggle.  It was an explicit attempt to change
project governance.   Some of the reasons some people wanted that
change were, in fact, related to commercial efforts (AFAICT).

There's no obvious good guys or bad guys in that history -- but it's
_interesting_.  It was a major development in the history of free
software.  It's important -- and from an outsider perspective, I'd
guess that everyone involved is now sufficiently distant from the
events to look at them pretty objectively.

By all means, edit the FAQ.  But please -- don't paint over history.
Perhaps this:  in the FAQ, the answer to this question:

  <h2><a name="cygnus">What is the relationship between GCC and Cygnus
  / Red Hat?</a></h2>

might be something like:

  At one time, that was a very controversial question.   It is 
  less controversial now.   Nevertheless, the events that raised the
  question have some historic interest for the free software movement.
  Some of this history has been recorded in "./EGCS-history".


And in "./EGCS-history" -- how about a paragraph or two from the
various people who were involved in egcs and the formation of the
steering commitee.   A first hand account from a slightly removed
perspective.   I'd give 40% odds that future historians will thank you
for your forsight.   Please consider inviting RMS to contribute a
section, if you go this route.

And anyway, such an attempt at objectivity and sober evaluation of
"how free software works" would be a lot more interesting (or at least
easier to interpret) than, say, jwz's account of the Lucent Emacs/GNU
Emacs schism and "why it is impossible to work with RMS".  No?

-t

Re: What is the relationship between GCC and Cygnus / Red Hat?

[Tom Lord]

Sorry to trouble you with one more message -- I know how welcome my
contributions are here, but I wanted to expand briefly on two points:

	> There's no obvious good guys or bad guys in that history --
	> but it's _interesting_.  It was a major development in the
	> history of free software.

My perception is that, post the egcs->steering committee transition,
the idea of creating NPOs for project governance of major FOSS
projects has become popular.   I think that the egcs->sc history
helped to bring that about (even though the sc is not itself an NPO).

	> Some of the reasons some people wanted that change were, in
	> fact, related to commercial efforts (AFAICT).

Looking at people's email addresses, it appears that a substantial
subset of the Cygnus GCC hackers have moved on to other companies
where they continue to work on GCC.  I speculate that the shift away
from old-style FSF project management to the sc helped to enable that
mobility.

-t

Re: What is the relationship between GCC and Cygnus / Red Hat?

[Gerald Pfeifer]

On Fri, 9 May 2003, Russ Allbery wrote:
> I certainly have no objections to removing it entirely; GCC and in general
> GNU toolchain development hasn't looked even remotely dominated by Cygnus
> or Red Hat to me for a long, long time now.  But if that isn't acceptable
> for whatever reason, an alternate suggestion:

Thanks for the alternate suggestion!

Taking into account the original feedback, and after waiting another week
for further feedback, I applied I my original patch:

  http://gcc.gnu.org/ml/gcc/2003-05/msg00894.html

Joe's point that this is in fact not a FAQ certainly helped convince me
that this is indeed the right approach.  If we want to _thank_ specific
companies, that could/should go into gcc/doc/contrib.texi, I'd say.

> Remove these two paragraphs.  I think they come across as alarmist
> (unnecessarily so), since this is no longer a common conception (at least
> as far as I'm concerned).

Absolutely, yes.

Gerald
-- 
Gerald "Jerry"   pfeifer@dbai.tuwien.ac.at   http://www.pfeifer.com/gerald/

Re: What is the relationship between GCC and Cygnus / Red Hat?

[Gerald Pfeifer]

On Fri, 9 May 2003, Tom Lord wrote:
> There's no obvious good guys or bad guys in that history -- but it's
> _interesting_.  It was a major development in the history of free
> software.  It's important -- and from an outsider perspective, I'd
> guess that everyone involved is now sufficiently distant from the
> events to look at them pretty objectively.
> [...]
> And in "./EGCS-history" -- how about a paragraph or two from the
> various people who were involved in egcs and the formation of the
> steering commitee.   A first hand account from a slightly removed
> perspective.   I'd give 40% odds that future historians will thank you
> for your forsight.   Please consider inviting RMS to contribute a
> section, if you go this route.

I fully agree that a historical description of the history of gcc/EGCS/GCC
would be very interesting from several points of view, but I'm afraid it's
quite hard to do that well (from a scientific point of view):

There aren't too many who have in-depth knowledge of the gcc 2 era, and of
those I suppose many will have a very subjective perspective.  Some of the
relevant information is not public (the old gcc2 list, private mails before
the formation of egcs, egcs and GCC SC mails, especially discussions with
RMS) and providing a fair historical overview will be highly non-trivial
(and even then it's not clear whether it will be accepted for inclusion in
official GCC documention by all affected parties).

Gerald
-- 
Gerald "Jerry"   pfeifer@dbai.tuwien.ac.at   http://www.pfeifer.com/gerald/

Re: What is the relationship between GCC and Cygnus / Red Hat?

[Joe Buck]

On Fri, 9 May 2003, Tom Lord wrote:
> > There's no obvious good guys or bad guys in that history -- but it's
> > _interesting_.  It was a major development in the history of free
> > software.  It's important -- and from an outsider perspective, I'd
> > guess that everyone involved is now sufficiently distant from the
> > events to look at them pretty objectively.
 
On Sat, May 17, 2003 at 12:16:57PM +0200, Gerald Pfeifer wrote:
> I fully agree that a historical description of the history of gcc/EGCS/GCC
> would be very interesting from several points of view, but I'm afraid it's
> quite hard to do that well (from a scientific point of view):
 
> There aren't too many who have in-depth knowledge of the gcc 2 era, and of
> those I suppose many will have a very subjective perspective.

I know the history, but as a participant my perspective would be highly biased.
In any case, any realistic discussion would re-open a lot of old wounds, wounds
that we spent a lot of effort patching together (it took about a year of delicate
discussion to repair the egcs-GCC rift, for example).

Re: What is the relationship between GCC and Cygnus / Red Hat?

[Mike Stump]

On Saturday, May 17, 2003, at 03:16 AM, Gerald Pfeifer wrote:
> There aren't too many who have in-depth knowledge of the gcc 2 era

Speak for yourself.  2.0 was the new gcc compiler.

:-)