This is the mail archive of the
mailing list for the GCC project.
Re: [Proposed binutils PATCH] Re: Diagnosing an intricate C++ pr
- To: "H . J . Lu" <hjl at lucon dot org>
- Subject: Re: [Proposed binutils PATCH] Re: Diagnosing an intricate C++ pr
- From: "Alexander N. Kabaev" <ak03 at gte dot com>
- Date: Tue, 05 Sep 2000 15:31:04 -0400 (EDT)
- Cc: binutils at sources dot redhat dot com, gcc-patches at gcc dot gnu dot org, "David O'Brien" <obrien at NUXI dot com>
- Organization: GTE Laboratories Inc.
> No. YOU have to make sure HEAD is ok for YOU.
I tried to explain why HEAD is the poor choice most of the time. YOU have
ignored that explanation. HEAD just cannot be in OK state all the time.
Are you suggesting all Linux users should stop running stable kernels and start
making sure development kernel is OK for THEM? What about KDE developers for
instance? Don't you think they have some other things to spend time on?
> binutils is used by many platforms. Checking things into the release
> branch needs to test on all those supported platforms. I simply don't
> have the time to check my patches into 2 trees. Also sometimes it is
> hard to tell a new feature from a bug fix. For some people, it may be
> a bug fix, but for others it may be a new feature.
When it is hard to tell what category given patch belongs to, patch should
remain on the HEAD branch only. It can be backported later if bug it fixes is
deemed too serious to ignore.
> There are rooms for improvement. But someone has to spend time to do it.
> If you want the release branch being updated, you should join the binutils
> development and do it.
For such an undertaking to succeed, it is important that binutils developers
at least realize the importance of proper release management. Reading your
posts one can conclude that you surely do not. Hopefully, other binutils
developers do not share your position. Release maintanance just cannot be done
without cooperation from the developers.
> I can only speak for myself. I don't have the
> time to spend on the release branch.
In other words, you officially abandon your users immediately after release has
been cut off.
There is no need to continue this discussion as it is unlikely that something
useful will come out of it. One more time, I just tried to express how the
current situation with binutils releases looks for an outside observer. And I
happen to think that I am not alone to have such thoughts.