Re: [PATCH, RFC] Enable IBM long double for PPC32 Linux

[David Edelsohn <dje at watson dot ibm dot com>]

>>>>> Giovanni Bajo writes:

Giovanni> This is a little unfair, though. So now the burden on enforcing the policy is
Giovanni> not on the maintainers that prepare the patches? The people involved in this
Giovanni> change have been working on GCC much longer than those who (later) objected.
Giovanni> They should have known our rules much better, and they should have asked a
Giovanni> buy-in from SC before starting this work, instead of silently forcing it in,
Giovanni> and then see if they could shut up the people who object (if any).

	This is an unfair characterization.  Target-specific changes have
been committed to GCC close to a release ever since I started working on
GCC over fifteen years ago.  Every Release Manager has tried to accomodate
port maintainers.

	I also do not see anyone trying to prevent people from objecting.
I do see a few people repeatedly raising the same objections without
constructive suggestions, despite public replies responding to the
concerns.  That type of discussion is not productive.

	I am sorry that you and others are upset.  I think everyone would
have appreciated greater communication and earlier notice.  That does not
change the current situation.  I do not see any reason to penalize the
targets who are trying to work within the current constraints to achieve
functional GCC and GLIBC releases.

Giovanni> I won't buy the argument "I won't hold up the release for this" as well, since
Giovanni> it misses the point that many important resources in GCC are being used in
Giovanni> fixing and testing this new feature, instead of putting GCC in shape for the
Giovanni> release. So the release has been already delayed because of this, and will be
Giovanni> even more. That's something which already happened.

	You seem to be redirecting your frustration about the release onto
the long-double-128 work as a scapegoat.  Neither you nor the GCC Release
Manager nor the GCC SC can force any GCC developer to work on any
particular project.  If the resources were not directed at
long-double-128, there is no guarantee that they would be directed as the
GCC 4.1 release or bugfixes or patch review or any other particular
project that you want.  I am sorry that your particular priorities are not
progressing as quickly as you wish, but lashing out at other developers
and their work will not solve that problem.

David