[aur-general] Unexpected aur package removal.

[Yichao Yu]

On Wed, Mar 13, 2013 at 12:49 AM, Daniel Wallace
<danielwallace at gtmanfred.com> wrote:
> On Wed, Mar 13, 2013 at 12:41:50AM -0400, Yichao Yu wrote:
>> On Wed, Mar 13, 2013 at 12:34 AM, Daniel Wallace
>> <danielwallace at gtmanfred.com> wrote:
>> > On Wed, Mar 13, 2013 at 12:17:13AM -0400, Yichao Yu wrote:
>> >> On Tue, Mar 12, 2013 at 11:49 PM, Daniel Wallace
>> >> <danielwallace at gtmanfred.com> wrote:
>> >> > On Tue, Mar 12, 2013 at 11:25:58PM -0400, Yichao Yu wrote:
>> >> >> Hi,
>> >> >>
>> >> >> A TU "gtmanfred" have just decided to remove one of my package just
>> >> >> one minute after he commented on the package pointed out the missing
>> >> >> package() in the PKGBUILD without any further explaination.
>> >> >>
>> >> >> The package is python-django-git[0], which I have uploaded 2-3 days
>> >> >> ago[1] to replacing two old packages including one with 3 vote that I
>> >> >> uploaded last Nov.
>> >> >>
>> >> >> I have already re-upload the package[0] since I don't think anything
>> >> >> is wrong with the package (especially not for the package name, I can
>> >> >> fix it if anything else with the package is wrong). And I just want to
>> >> >> know if that was a delete by mistake or a missing package() somehow is
>> >> >> enough reason to remove a package within ~1min after notifying the
>> >> >> maintainer now?
>> >> >>
>> >> >> Yichao Yu
>> >> >>
>> >> >> [0] https://aur.archlinux.org/packages/python-django-git/
>> >> >> [1] https://mailman.archlinux.org/pipermail/aur-general/2013-March/022494.html
>> >> >
>> >> > Please follow packageing guidelines, anything that touches $pkgdir
>> >> > should be inside the package() function.
>> >> >
>> >> > repackaging without a package() function has been deprecated
>> >> > https://mailman.archlinux.org/pipermail/pacman-dev/2010-April/010620.html
>> >> >
>> >> > and PKGBUILDs without a package() function have been depricated
>> >> > https://patchwork.archlinux.org/patch/515/
>> >> > https://www.archlinux.org/todo/clean-up-pkgdir-usage/
>> >>
>> >> I agree, and I have already fixed it.
>> >>
>> >> BUT, that is NOT what I am asking!! What I am REALLY asking here is
>> >> why did you simply remove the package!!
>> >> You can leave a comment (which you did 1min before you go right to the
>> >> incorrect last step), flag it out-of-date, send me a email, or even
>> >> disown it and correct it yourself according to the AUR two-week policy
>> >> if I refuse to update. Is it what you think a TU should do to remove
>> >> (without waiting for the shortest reasonable response time or even
>> >> attempting to improve) all non-standard/old PKGBUILD on AUR.
>> >>
>> >> It is fine if you have just removed it by accident (although I will
>> >> probably suggest to move the remove button and the flag-out-of-date
>> >> button farther away for TU if that's the case), but if you were doing
>> >> that on purpose, what you did is totally non-constructive. This is
>> >> definitely NOT what a TU should do. Let me remind you what a TU should
>> >> do is "check PKGBUILDs for minor mistakes, suggest corrections and
>> >> improvements"[1], I am not sure which of the three does "removing
>> >> package right away" belong to.
>> >>
>> >> [1] https://wiki.archlinux.org/index.php/AUR_Trusted_User_Guidelines#The_TU_and_.5Bunsupported.5D
>> >>
>> >> >
>> >> > --
>> >> > Daniel Wallace
>> >> > Archlinux Trusted User (gtmanfred)
>> >> > Georgia Institute of Technology
>> >
>> > If you read the front page of the AUR, you will note the following.
>> >
>> > Contributed PKGBUILDs must conform to the Arch Packaging Standards
>> > otherwise they will be deleted!
>>
>> OK, as you are happily deleting every single slightly non-standard AUR
>> packages, I want to ask on the mailinglist when did removing package
>> immediately becoming the rule.
>>
>> >
>> > https://aur.archlinux.org/
>> >
>> > --
>> > Daniel Wallace
>> > Archlinux Trusted User (gtmanfred)
>> > Georgia Institute of Technology
>
> For the record, I have only deleted NEWLY updated packages that are
> either.
>
> A) Named incorrectly.
>         Someone uploads a -git or -svn package and doesn't name it -git or
>         -svn, it has no votes and no comments, so instead of waiting for it
>         to get to a point where it would have to be merged, I leave a note
>         explaining that it should be named correctly and then delete it.

This is probably fine but I think there should REALLY be a guide line
on wiki or AT LEAST a NOTE on the mailing list ~1-2days before you
start to do that.

>
> B) it has no package() function.
>         I am not going through all of the aur and deleting everything
>         without a package() function.  I am deleting newly uploaded packages
>         that don't have a package() function.  I noticed yours had 3 votes,
>         I assumed it was because it was one of the first ones in my rss feed
>         of new packages to the aur, and one of the oldest since the last
>         time I had gone through them.  If you check the AUR stats on teh
>         side of the home page, you will not that 1/4th of all packages in
>         the aur have never been updated.  So instead of just leaving a note
>         to add a package() function I delete them because it is brand new,
>         almost no votes and usually no comments.

I'm assuming you remove "brand" new packages because they have no user. But
1, this is a simple rename of an old package, that is NOT NEW
2, there are 3 votes, which means it has user
3, since you already has rss for packages, why can't you be more
responsible (as a TU) to come back at the package at least a couple of
days later.

>
>         On this point, I don't believe we should be merging to any packages
>         which do not properly meet the Packaging Guidelines.
>
>
> --
> Daniel Wallace
> Archlinux Trusted User (gtmanfred)
> Georgia Institute of Technology