[forum] Suggestion for XFree86

David Dawes forum@XFree86.Org
Thu, 20 Mar 2003 17:31:26 -0500

On Thu, Mar 20, 2003 at 02:02:16PM -0800, Eric Anholt wrote:
>On Thu, 2003-03-20 at 13:10, David Dawes wrote:
>> On Thu, Mar 20, 2003 at 12:31:58PM -0800, Eric Anholt wrote:
>> >On Thu, 2003-03-20 at 11:59, David Dawes wrote:
>> >> On Thu, Mar 20, 2003 at 01:34:25PM -0500, Alan Cox wrote:
>> >> >> depends on XFree86's continued success.  David Dawes has turned to that
>> >> >> constituency to assist him in resolving this situation.  Both of these
>> >> >
>> >> >After kicking Keith out. Bad mistake but whats done is done.
>> >> >
>> >> >> The ONLY thing XFree86 needs to do to correct this problem is to establish a
>> >> >> mechanism where any interested person may join XFree86 and be given an equal
>> >> >> vote to elect a slate of candidates for a governing board of directors whose
>> >> >
>> >> >This isnt about voting, governance and dictatorship IMHO, its about
>> >> >change. XFree86 is hard to get involved with usefully, resistant to cool
>> >> >ideas and strongly wedded to an occasional not rolling regular release
>> >> >model.
>> >> 
>> >> Keith already put a lot of cool stuff in, and had a very free hand in
>> >> doing so.  That he did this isn't tied to him being a core team member.
>> >> Others do cool stuff to.  XFree86 rejects very little of what is submitted.
>> >
>> >What exactly is Keith's status as a committer?  I had the impression
>> >from talking to others that he was no longer able to commit to XFree86.
>> Yes, that is correct, and that has been true since he abused his
>> commit privilege by committing XFIXES on the eve of the 4.3 freeze
>> without any prior discussion about it.  It didn't prevent him from
>> continuing to contribute to the 4.3 release after that point, as
>> can be seen from the CHANGELOG.
>How was that decision made?  Committing changes before a freeze sounds
>fine to me.  If it lacked review, it should have just been backed out
>(preferably by the original committer).

Adding a new "standard" extension (it was documented as such)
without any prior discussion, let alone public review, is not simply
"committing changes".  With such a long release cycle, there was
plenty of time to get these things done sooner.  After all, they
are claimed to be fixes for very long standing flaws in the X11
core protocol that the proponents have been aware of for years.
Why wait until the eve of the freeze?

I've heard rumours that the timing was actually a setup to provoke
a crisis, but you have to be careful about rumours...

David Dawes
Release Engineer/Architect                      The XFree86 Project