[gentoo-dev] metadata revised - removal of packages

[gentoo-dev] metadata revised - removal of packages

[Torsten Veller]

1)
While removing maintainers from metadata i realised (not for the first
time) that i don't understand metadata :/

I thought it might be a good idea to add a new herd (maintainer-needed)
for packages where the maintainer has left. I think it will make it
    easier for bugwranglers to assign bugs
    transparent that the maintainance situation has changed
    trackable

To me maintainer-needed is always a first step before a package gets
masked and removed from the tree.

Well, i was told that adding the maintainer-needed herd is not a good
idea and it is best to remove metadata.xml if no valuable information
remains.

I couldn't find information on that. Can somebody explain?



2)
What is the next step after the last maintainer is removed from metadata.xml?
Well i announced these packages on -dev. Now i can wait some time (how
long?) and then?
-- 
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] metadata revised - removal of packages

[Jakub Moc]

[-- Attachment #1: Type: text/plain, Size: 1111 bytes --]


9.9.2005, 11:58:21, Torsten Veller wrote:


> Well, i was told that adding the maintainer-needed herd is not a good
> idea and it is best to remove metadata.xml if no valuable information
> remains.

> I couldn't find information on that. Can somebody explain?

I don't understand this idea on removing metadata.xml. There are lots of
packages with bug-wranglers@gentoo.org as herd in metadata. Removing
metadata.xml is IMHO a really bad idea. Just will make me look at the
ChangeLog, assign to maintainer-needed and CC someone who has touched the
ebuild most often.

> 2)
> What is the next step after the last maintainer is removed from metadata.xml?
> Well i announced these packages on -dev. Now i can wait some time (how
> long?) and then?

Put maintainer-needed in herd seems a logical solution to me. Hmmm. Maybe I
missed something?

-- 
Best regards,

 Jakub Moc
 mailto:jakub@gentoo.org
 GPG signature: http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xCEBA3D9E
 Primary key fingerprint: D2D7 933C 9BA1 C95B 2C95  B30F 8717 D5FD CEBA 3D9E

 ... still no signature ;)

[-- Attachment #2: Type: application/pgp-signature, Size: 183 bytes --]

Re: [gentoo-dev] metadata revised - removal of packages

[Mike Frysinger]

On Friday 09 September 2005 05:58 am, Torsten Veller wrote:
> 2)
> What is the next step after the last maintainer is removed from
> metadata.xml? Well i announced these packages on -dev. Now i can wait some
> time (how long?) and then?

and then what ?  if you're proposing removal of packages due solely to no 
maintainer, then we're going to have to slap you around.  dont remove 
packages for that reason alone.
-mike
-- 
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] metadata revised - removal of packages

[Alec Warner]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Mike Frysinger wrote:
> On Friday 09 September 2005 05:58 am, Torsten Veller wrote:
> 
>>2)
>>What is the next step after the last maintainer is removed from
>>metadata.xml? Well i announced these packages on -dev. Now i can wait some
>>time (how long?) and then?
> 
> 
> and then what ?  if you're proposing removal of packages due solely to no 
> maintainer, then we're going to have to slap you around.  dont remove 
> packages for that reason alone.
> -mike

So I guess the idea would then be, how do you find packages in the tree
where dev FooGuy once maintained it but no longer does (because FooGuy
left) and the package is old and nasty and no one cares about it.  Leave
it in the tree anyway?  I certainly don't want unmaintained CRAP in the
tree, although unmaintained decent programs are good.  By decent I mean
programs that are generally so old they never have version bumps ;)

- -Alec
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.1 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org
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=vQKJ
-----END PGP SIGNATURE-----
-- 
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] metadata revised - removal of packages

[Chris Gianelloni]

[-- Attachment #1: Type: text/plain, Size: 1487 bytes --]

On Fri, 2005-09-09 at 09:29 -0400, Alec Warner wrote:
> > and then what ?  if you're proposing removal of packages due solely to no 
> > maintainer, then we're going to have to slap you around.  dont remove 
> > packages for that reason alone.
> > -mike
> 
> So I guess the idea would then be, how do you find packages in the tree
> where dev FooGuy once maintained it but no longer does (because FooGuy
> left) and the package is old and nasty and no one cares about it.  Leave
> it in the tree anyway?  I certainly don't want unmaintained CRAP in the
> tree, although unmaintained decent programs are good.  By decent I mean
> programs that are generally so old they never have version bumps ;)

The simple rule is leave it the hell alone.  If it is completely broken,
there will be bugs filed.  If nobody steps up to claim the package,
then, and only then, is it removed.

Did you look at the list of packages?  I mean, how exactly broken can
"descent1-maps" get?  *grin*

Package removal because it has no maintainer would probably remove a
large portion of the tree, possibly even packages that are necessary for
many people.  There's quite a few packages that get maintained simply by
people fixing problems with them, but with no real "maintainer".  While
this isn't the best solution, removing them from the tree just for this
reason is asinine.

-- 
Chris Gianelloni
Release Engineering - Strategic Lead/QA Manager
Games - Developer
Gentoo Linux

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

Re: [gentoo-dev] metadata revised - removal of packages

[Diego 'Flameeyes' Pettenò]

[-- Attachment #1: Type: text/plain, Size: 1148 bytes --]

On Friday 09 September 2005 14:46, Mike Frysinger wrote:
> and then what ?  if you're proposing removal of packages due solely to no
> maintainer, then we're going to have to slap you around.  dont remove
> packages for that reason alone.
Exactly the point. And I follow this request. If that was the case, publib and 
libiconv would have been removed from the tree a lot of time ago, while they 
are now maintained by BSD herd and by me, as they are useful for portability.

Unfortunately, seems like Ciaran thinks that removing them is the only way to 
go, also if they can be perfect and just be unmaintained, because ebuilds 
risk to not pass future repoman tests and needs to be qa-checked from time to 
time, adding more work.

Well that's why we have a QA team, isn't it? For most of the changes in 
repoman check is possible to run a whole-tree check and then who propose the 
change can take care of fixing the unmaintained parts, like I did for 
enewuser/cp-a/chown root:root bugs.

-- 
Diego "Flameeyes" Pettenò
Gentoo Developer - http://dev.gentoo.org/~flameeyes/
(Gentoo/FreeBSD, Video, Gentoo/AMD64, Sound, PAM)

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

Re: [gentoo-dev] metadata revised - removal of packages

[Ciaran McCreesh]

[-- Attachment #1: Type: text/plain, Size: 559 bytes --]

On Fri, 9 Sep 2005 17:23:00 +0200 "Diego 'Flameeyes' Pettenò"
<flameeyes@gentoo.org> wrote:
| Unfortunately, seems like Ciaran thinks that removing them is the
| only way to go

Uh, no, I said that the ideal thing to do would be to find a new
maintainer, or failing that remove the package. I didn't say that
critical things should be removed the instant the maintainer vanishes.

-- 
Ciaran McCreesh : Gentoo Developer (Vim, Shell tools, Fluxbox, Cron)
Mail            : ciaranm at gentoo.org
Web             : http://dev.gentoo.org/~ciaranm


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

Re: [gentoo-dev] metadata revised - removal of packages

[Chris Gianelloni]

[-- Attachment #1: Type: text/plain, Size: 1165 bytes --]

On Fri, 2005-09-09 at 11:58 +0200, Torsten Veller wrote:
> To me maintainer-needed is always a first step before a package gets
> masked and removed from the tree.

Agreed.  Set the herd to maintainer-needed.  Since there isn't an actual
maintainer-needed herd, there's no need to add it to herds.xml or
anything.

> Well, i was told that adding the maintainer-needed herd is not a good
> idea and it is best to remove metadata.xml if no valuable information
> remains.

I would say definitely not.  The herd information can be polled by
jeeves, so it *is* valuable to have maintainer-needed in there.

> I couldn't find information on that. Can somebody explain?

Everything should have metadata.  Removing it only adds more work for
someone else later.

> 2)
> What is the next step after the last maintainer is removed from metadata.xml?

Reassign all bugs open for the package to maintainer-needed.

> Well i announced these packages on -dev. Now i can wait some time (how
> long?) and then?

Nothing.  You leave them alone.

-- 
Chris Gianelloni
Release Engineering - Strategic Lead/QA Manager
Games - Developer
Gentoo Linux

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

Re: [gentoo-dev] metadata revised - removal of packages

[Ciaran McCreesh]

[-- Attachment #1: Type: text/plain, Size: 910 bytes --]

On Fri, 9 Sep 2005 11:58:21 +0200 Torsten Veller <tove@gentoo.org>
wrote:
| Well, i was told that adding the maintainer-needed herd is not a good
| idea and it is best to remove metadata.xml if no valuable information
| remains.
| 
| I couldn't find information on that. Can somebody explain?

maintainer-needed isn't a herd, it's an alias. On the same subject,
no-herd isn't legal in metadata.xml either. Both of these are
metastructure policy matters, so changing them requires a GLEP.

A GLEP probably wouldn't be a bad idea, either... The way the tree is
maintained has changed considerably over the past few years. In
particular, herds are no longer a fallback for when the maintainer
(single person) is not available.

-- 
Ciaran McCreesh : Gentoo Developer (Vim, Shell tools, Fluxbox, Cron)
Mail            : ciaranm at gentoo.org
Web             : http://dev.gentoo.org/~ciaranm


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

Re: [gentoo-dev] metadata revised - removal of packages

[Mike Frysinger]

On Friday 09 September 2005 12:10 pm, Ciaran McCreesh wrote:
> On Fri, 9 Sep 2005 11:58:21 +0200 Torsten Veller <tove@gentoo.org>
> | Well, i was told that adding the maintainer-needed herd is not a good
> | idea and it is best to remove metadata.xml if no valuable information
> | remains.
> |
> | I couldn't find information on that. Can somebody explain?
>
> In
> particular, herds are no longer a fallback for when the maintainer
> (single person) is not available.

that is policy left up to each herd imo ... rarely do we list a maintainer in 
a base-system or games packages
-mike
-- 
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] metadata revised - removal of packages

[Chris Gianelloni]

[-- Attachment #1: Type: text/plain, Size: 435 bytes --]

On Fri, 2005-09-09 at 17:10 +0100, Ciaran McCreesh wrote:
> maintained has changed considerably over the past few years. In
> particular, herds are no longer a fallback for when the maintainer
> (single person) is not available.

Say what?  What the hell is the point of a herd, then?  And when did
this nonsense happen?

-- 
Chris Gianelloni
Release Engineering - Strategic Lead/QA Manager
Games - Developer
Gentoo Linux

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

Re: [gentoo-dev] metadata revised - removal of packages

[Ciaran McCreesh]

[-- Attachment #1: Type: text/plain, Size: 1076 bytes --]

On Fri, 09 Sep 2005 13:49:55 -0400 Chris Gianelloni
<wolf31o2@gentoo.org> wrote:
| On Fri, 2005-09-09 at 17:10 +0100, Ciaran McCreesh wrote:
| > maintained has changed considerably over the past few years. In
| > particular, herds are no longer a fallback for when the maintainer
| > (single person) is not available.
| 
| Say what?  What the hell is the point of a herd, then?  And when did
| this nonsense happen?

This is what the original metastructure policy said, which means it
happened something like three years ago. The original point of herds
was as a fallback rather than as a primary maintainer. This obviously
isn't how things are done currently... A GLEP revising the role of
herds and adding in maintainer-needed would be useful. I'm not going to
be around over the weekend, but if no-one else has written something up
by whenever I get back I'll throw a draft together...

-- 
Ciaran McCreesh : Gentoo Developer (Vim, Shell tools, Fluxbox, Cron)
Mail            : ciaranm at gentoo.org
Web             : http://dev.gentoo.org/~ciaranm


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