Re: [gentoo-dev] Re: [gentoo-commits] gentoo-x86 commit in dev-python/PyZilla: PyZilla-0.1.0.ebuild ChangeLog metadata.xml

[Rich Freeman]

On Sun, Mar 27, 2011 at 9:30 AM, Jeremy Olexa <darkside@g.o> wrote:
> On 03/27/2011 02:47 AM, Nirbheek Chauhan wrote:
>>> On 03/26/2011 12:52 AM, Mike Frysinger (vapier) wrote:
>> I propose that we should be more aggressive about package.masking (for
>> removal) all maintainer-needed packages from the tree by doing that
>> one month after they become maintainer-needed. If someone doesn't
>> volunteer to take care of it, it probably wasn't important anyway.
>>
>
> That is abit extreme for me (read: I don't have motivation to fight the
> flames), but I wouldn't complain if someone else did it to be honest.
>

So, I'd like to propose that somewhere between adding stuff to the
tree that nobody has any intent to look after, and removing stuff that
has been around a long time with no clear problems, there is a happy
medium.

How about this - if you add a package to the tree, you are responsible
for it for at least a year.  If you can get somebody else to take it
then that is fine.  If it has problems QA can flame you (privately at
first) for it, and you should feel appropriately embarrassed and fix
it, or remove it.

After a year, it can go maintainer-needed.  Before a year, it cannot,
and you either need to actually maintain it, or remove it.  Developers
should not be adding packages they have no interest in whatsoever, or
that have so many QA issues initially that they're high-maintenance
right from the start.  If a dev gets a package from a proxy-maintainer
and they disappear then they can nurse it along or remove it as makes
sense - we should be nice to these devs but we shouldn't just cut the
packages loose.

Packages that are maintainer-needed stay around as long as they're not
making trouble.  If they get lots of complaints they get announced on
-dev, and after two weeks they get masked if not picked up.  If they
end up blocking something then likewise they get announced and then
masked.  That basically is the current practice anyway.

I don't see a need to remove m-n packages wholesale just to say that
we did it, or to punish users for not becoming devs or whatever.

And of course, the usual long-term solutions like making
proxy-maintaining easier should be pursued.

Rich