Request For Candidates: Application Review Board
Stephan Hermann
sh at sourcecode.de
Mon Aug 16 08:38:10 BST 2010
Moins,
Rick,
On Friday, August 13, 2010 10:49:22 pm Rick Spencer wrote:
> On Fri, 2010-08-13 at 16:34 -0400, Scott Kitterman wrote:
> > > This is awesome.
> >
> > This definitely something. That's not the word I would pick.
>
> Ok, you've expressed your opinion, and I thank you for that. I can see
> you won't be volunteering for the application review board, which is
> fine as well.
>
> As the specs were written and the blueprints have been largely
> implemented, I don't see anything particularly likely to change,
> including your view on the matter, so I suppose this particular thread
> has come to a conclusion.
This was actually a liver punch and regarding the reputation and the work of
Scott for Ubuntu that was really bad. With such an attitude contributors are
pushed away.
Despite the fact, that some blueprints and specs are written, and I do think I
understand why "we" want this to happen, I do agree with Scott and the other
fellow MOTUs, that this way of pushing apps into Ubuntu is a bit awkward.
First, in the beginning we wanted to have every and any FLOSS software which
is available inside Ubuntu (universe/multiverse).
We encouraged every and any app developer if they want to see their apps in
Ubuntu, that they go the REVU way. When they didn't have the right packaging
skills we asked upstream to find people who do have the right packaging skills.
Now, REVU lacks manpower, it's not a nice work to do, because of people not
responding, or you get packages which are quite good, we upload, and the
"maintainer" is just gone and isn't interested in the package anymore.
Now, we are starting with a "different" archive (it has to be a different
archive, if not, then this proposal is even more broken for me) for app
developers who can push their "important" apps into ubuntu, without thinking
about the next day.
Is that really what we want?
If you want masses, you don't check for quality, you just push and push.
We want applications which are high quality, not poor quality. Having a
reviewing board in place, the process will first start like a thunderstorm and
then after a couple of weeks or months it will starve. Just like REVU.
And regarding commercial app developers who want to have their shareware or
whatever in our repositories, there is always the way to the canonical partner
repository.
So, which customers do you expect to use this process? Upstream developers
don't actually have any deep inside of a distro, that's why Debian, Fedora and
other distros do have package maintainers, which are mostly != upstream.
Do you really want to have amateure developers who are creating some quickly
apps or small python scripts, or do you want to have software which has a
future?
If you want to have software in our archive which has a future, we should
stick with the old process, go the REVU way, package your app nicely and with
a good quality, and have it permanently in Ubuntu (universe/mutliverse).
Furthermore, how do we track bugs for those apps? Inside project ubuntu? Or
separated from the normal ubuntu bugtracker?
Anyways, Pitti already wrote what this is all about:
Quote of Pittis mail from Date: Sun, 15 Aug 2010 13:12:09 +0200:
"
Hello Scott,
Scott Kitterman [2010-08-13 16:34 -0400]:
> That's nonsense. With backports you can have stuff in the development
release
> and in backports on the same day.
But we don't want all those third-party apps in Ubuntu, since we would
earn the responsibility of maintaining them. That's the main point of
the new processes, as far as I can see. We want to make it easier to
discover third-party apps, but not "swallow" them all.
Martin
"
If this process is about what Pitti wrote, we can still stick with LP PPAs and
software center could provide a list with all PPAs available from LP. This is
an easy task with LP lib I think.
Then we have something like the OpenSuSE BuildService with hundreds of
different packages...we don't provide support for them (if you upload to an
Ubuntu maintained archive, Ubuntu needs to provide support for those apps,
even if nobody thinks about that, but users are expecting it).
Before I would vote for suche a process, I would really like to have a little
inside (Ubuntu wise and also Canonical wise) which customer base you really
want to serve with this new process.
Regards,
\sh
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 316 bytes
Desc: This is a digitally signed message part.
Url : https://lists.ubuntu.com/archives/ubuntu-devel/attachments/20100816/5f4c2f4a/attachment.pgp
More information about the ubuntu-devel
mailing list