Re: NMU ultimatum within less then 12h is rude (was: Bug#118388: intent to NMU merlin-cpufire)

[Sam Hartman <hartmans@debian.org>]

>>>>> "Bernd" == Bernd Eckenfels <ecki@lina.inka.de> writes:

    Bernd> On Tue, Nov 06, 2001 at 05:56:18PM -0500, Patrick Ouellette
    Bernd> wrote:
    >> Lighten up (both of you) - Bernd has a point, doesn't want
    >> someone creating a mess he has to clean up later.  Branden has
    >> just as valid a point - he is BEING PAID to work on ia64
    >> compatibility of packages, and can devote time to fixing build
    >> problems on the ia64.  Why can't the two of you work *together*
    >> to solve the problem, instead of getting in a public pissing
    >> match that does no good for anyone (except suck up bandwidth)?

    Bernd> Because Branden is not the only one doing that, and my
    Bernd> point is much more deeper. If it is (anywhere) agreed to
    Bernd> have a 4-day idle limit for uncritical updates the debian
    Bernd> project is surely not for me anymore.

You know there are operating systems out there where all the
developers have full commit access and check in changes to areas of
code that are not strictly theirs on a daily basis.  Guess what: those
operating systems do not have clearly different observed quality for
their unstable releases.

I get the feeling that many developers feel that someone NMUing one of
their packages implies they did something wrong or are a bad
developer.  This is partially because in some cases it does.  But I am
very disappointed when it seems that a developer is objecting to an
NMU not because it will break things but because they see it as a
challenge to their control over their package.  Remember we are all
here to improve the quality of Debian.  You should not stand in the
way of a NMU that correctly fixes a problem; you should thank someone
for making your job easier.

Instead of getting pissed off just because someone did an NMU, why
don't you evaluate the situation.  First, did they break things.  If
so, then by all means let your fury know no bound, or at least pass
along the frustration the broken NMU has caused you.

Of course if the NMU is correct you may still be worried that it was
correct only by accident.  It seems reasonable to ask someone who did
an NMU of your package how they tested their change and made sure
their NMU is correct.  If they didn't follow the standards of care you
have established for your package, or if in the absence of explicit
standards for the package, they didn't follow reasonable standards of care, you have something to be up set about.  Probably you will be almost as up set as if a broken package had been uploaded.  

But if the NMU is correct and was prepared in a manner that maximizes
the likelyhood that the NMU would be correct, you don't have anything
to be angry about.   

That aside, under current policy/practice, I don't thing Branden has a
leg to stand on if he violates the wishes of the maintainer on NMUs.
IN many ways I think this is unfortunate, but there is a fairly strong
consensus on this.