[gentoo-dev] Fw: [gentoo-core] [POLICY] Keywording/Stabilizing Bug Assignment Policy

[gentoo-dev] Fw: [gentoo-core] [POLICY] Keywording/Stabilizing Bug Assignment Policy

[Jeroen Roovers]

Begin forwarded message:

Date: Wed, 18 Apr 2007 00:12:26 +0200
From: Jeroen Roovers <jer@gentoo.org>
To: gentoo-core@lists.gentoo.org
Subject: Re: [gentoo-core] [POLICY] Keywording/Stabilizing Bug
Assignment Policy


On Tue, 17 Apr 2007 21:50:26 +0200
"Stefan Schweizer" <genstef@gentoo.org> wrote:

> As a maintainer I have to deal with many stable/keywording requests.
> Those are bugs that generally hang around in my bugzilla queries and
> fill my mailbox and I do not have any ability to help there or fix
> them. Those bugmails constitute spam for my mailbox.

The bugs you open generate too much information? So basically you do
too much work to still cope with the consequences of that very same
work? Maybe you can get a dev to act as your secretary? (I am being
serious.)

This matter appears to be entirely unrelated to the original thread,
too.

> It would be cool to implement a keywording@gentoo.org alias just to
> assign those bugs to so that we maintainers do not need to see them.

So this way you would avoid receiving important information? Why don't
you set up bugsy not to inform you about these under [Email
Preferences]? Whether it is wise to do this, is a different matter
altogether.

> In my opinion the last architecture should also remove the old ebuild
> they have just made obsolete by stabling/keywording the new version,
> since they commit to the directory anyway.

Some more policy would be need on this second matter entirely unrelated
to the original thread. The last arch dev keywording a package should
then check as well whether the package:

1) is SLOTted.
2) has a history of having users choose specific versions (mask newer
versions, basically) based on their own needs.

If a package is not SLOTted, it is still not clear whether 2) is the
case, so IMHO only the package maintainer should ever (yet diligently)
clean up so-called "old" ebuilds.


Kind regards,
     JeR
-- 
gentoo-dev@gentoo.org mailing list

[gentoo-dev] Re: Fw: [gentoo-core] [POLICY] Keywording/Stabilizing Bug Assignment Policy

[Stefan Schweizer]

Jeroen Roovers wrote:
> On Tue, 17 Apr 2007 21:50:26 +0200
> "Stefan Schweizer" <genstef@gentoo.org> wrote:
>> As a maintainer I have to deal with many stable/keywording requests.
>> Those are bugs that generally hang around in my bugzilla queries and
>> fill my mailbox and I do not have any ability to help there or fix
>> them. Those bugmails constitute spam for my mailbox.
> 
> The bugs you open generate too much information? So basically you do
> too much work to still cope with the consequences of that very same
> work? Maybe you can get a dev to act as your secretary? (I am being
> serious.)

They generate irrelevant information for me because I cannot help with
stabling. So I would love to have the possibility to -CC me there after
siging the stabling off.

>> It would be cool to implement a keywording@gentoo.org alias just to
>> assign those bugs to so that we maintainers do not need to see them.
> 
> So this way you would avoid receiving important information?

The information is not important to me, the headlines are just filling my
mailbox

> Why don't 
> you set up bugsy not to inform you about these under [Email
> Preferences]?

This is possible? How can I separate keywording/stable and normal bugs in a
general way?

> [..] so IMHO only the package maintainer should ever (yet diligently)
> clean up so-called "old" ebuilds.

or the package maintainer can put a REMOVEOLD keyword on the bug or
something to make obvious that he wants it removed. I would love this
possibility :)

-- 
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Re: Fw: [gentoo-core] [POLICY] Keywording/Stabilizing Bug Assignment Policy

[Matti Bickel]

[-- Attachment #1: Type: text/plain, Size: 648 bytes --]

Stefan Schweizer <genstef@gentoo.org> wrote:
> They generate irrelevant information for me because I cannot help with
> stabling. So I would love to have the possibility to -CC me there after
> siging the stabling off.

You are responsible for any failures discovered by arch teams while
testing. Nothing sucks more than us discovering "ups, $maintainer hasn't
tested this code to be endian aware and it goes nuts on my machine" and
the maintainer not responding to it (yes, i'll prod $maintainer
seperatly on IRC, but bugs is just way easier).
-- 
Regards, Matti Bickel
Homepage: http://www.rateu.de
Encrypted/Signed Email preferred

[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]