[aur-general] AUR4 migration of orphan packages

[Ido Rosen]

I think some of the orphans on AUR are just maintained by multiple
people.  The usage pattern is:

Person A adopts, updates, and disowns.
Person B some time later notices it's out of date, adopts, updates, disowns.

It seems perfectly reasonable to have multiple people maintain a
package over time this way.  Maybe we just need better support for
this style of non-maintainership that isn't quite "orphaned"?  Support
for multiple maintainers/collaborators like on GitHub repos?
(Outright owning a package in AUR prevents anyone else from updating
it.)

I do something in between outright maintainership and this "adopt,
update, disown" non-maintainership: I have a git repo with my AUR
packages, and accept pull requests on GitHub -- if someone wants to
update a package faster than I can get to it (since I only have time
on weekends), they submit a pull request and I merge it in, test, and
submit to AUR (which takes 2 min to verify & submit the package, vs.
the a-priori-unknown time commitment of doing it all myself).  It
would be nice if there were an official way to make AUR support
collaborative maintainership like this.

On Tue, Jun 9, 2015 at 11:10 AM, Giancarlo Razzolini
<grazzolini at gmail.com> wrote:
> On 09-06-2015 08:17, Jesse McClure wrote:
>>
>> Agreed.  All the packages that no one carries over to aur4 will still be
>> archived for some time, so if anyone*actually*  wants them in aur4, they
>> can
>> adopt them.  One can keep their own store of PKGBUILDs, but the aur is for
>> packages that it is likely multiple users will want.  If not even*one*
>> person
>> wants a package enough to maintain it in the aur, then it doesn't need to
>> be
>> there.
>
> I have adopted some packages, created a few more, but I think that this
> migration should serve the purpose of cleansing the database. We already
> have orphans on aur4 and that is unacceptable. Migrate a package and then
> orphan it is not ideal and we will end up having the same number of orphans
> as we already have.
>
> Cheers,
> Giancarlo Razzolini