Bug#766708: Coordinating a plan and requirements for cross toolchain packages in Debian
Don Armstrong writes ("Bug#766708: Coordinating a plan and requirements for cross toolchain packages in Debian"):
> On Wed, 17 Dec 2014, Helmut Grohne wrote:
> > I also note that concern was brought for raising this issue to the
> > ctte too quickly. In fact, we now see that I was raising it too
> > slowly.
> #766708 was raised both too quickly (1 day between filing and
> reassigning), and too late (less than one month between the end of
> pre-freeze transitioning).
The bug was filed one day after the problematic upload. What you are
saying is that the upload was done too late. But that is precisely
one of the submitter's complaints!
None of this situation is the submitters' making.
> Unfortunately, the CTTE takes a very long time to figure things
IMO the committee has utterly failed to discharge its responsibility
here. If I had still been on the committee I would have pushed for a
swift vote to overrule.
> and it's not helped when one of the parties does not communicate.
That is also no excuse for not overruling a bad decision. Indeed, it
is allowing the maintainer to holding the process hostage by refusing
The only non-delegatable responsibilities of a Debian package
maintainer are (1) to communicate and (2) not to block other people's
work. The maintainer here has failed in both of those duties and the
TC has acquiesced:
The TC has stood by while a working feature was suddenly sabotaged at
the last minute before the freeze, for doubtful design purity reasons.
Frankly I'm impressed that Helmut and Wookey have managed to be as
calm and reasonable and friendly in the face of this situation, which
must be very frustrating for them.
Of course they are in a bind: if they argue too much, they may find
that they upset the GCC maintainer, and that things get even more
difficult for them. That is particularly troublesome because the TC
has shown that it will not step in when the maintainer's power is
Luckily my own work doesn't depend on cooperation from the GCC
maintainer, so I can say all of this.