Request For Candidates: Application Review Board

[Scott Kitterman]

On Monday, August 16, 2010 02:16:53 pm Jono Bacon wrote:
> Hi All,
> 
> Sorry for the delay in responding; I was on vacation and locked in a
> studio.
> 
> On Sun, 2010-08-15 at 13:08 +0200, Martin Pitt wrote:
> > Andrew SB [2010-08-13 15:23 -0400]:
> > > Who is responsible for pushing these new apps to the development
> > > release?
> > 
> > The author/packager of the software. AFAICS the main point of this
> > process was to avoid having to push a lot of "fire'n'forget" apps into
> > universe.
> > 
> > > As Scott said, "Once the next release hits they won't be in
> > > the current release anymore because we skipped the part where we get
> > > the packages into the development release." Are we creating a
> > > permanent split in the archive with two completely different
> > > processes?
> > 
> > Those new packages should not be considered being part of Ubuntu, the
> > new process should just provide a standard way of making third-party
> > apps more easy to discover. But the responsibility for those should
> > still by and large remain at the third party's side. Of course by
> > showing them in software-center we gain a bit of responsibility, which
> > is why we need to introduce a review step here.
> > 
> > Not sure how you think about this, but I think from a number of
> > maintainers/MOTUs POV just about the last thing universe needs is even
> > more undermaintained packages?
> 
> I think Martin hits the nail on the head here; the goal of this process
> is to raise the visibility of applications, and this may include some
> applications that don't have the steam to be fully maintained in
> Universe.
> 
> I think many of us agree that our current developer processes are simply
> *too heavy* for application authors who want to get their apps
> visibility in Ubuntu Software Center -- as one such example, I wrote
> Lernid and if I didn't know didrocks who could help get it packaged, I
> would have never have been able to get Lernid visible in Ubuntu Software
> Center without engaging in a set of packaging assessment processes that
> are too heavy for me. They are too heavy because our processes are
> design for people who build Operating Systems - as an app author I don't
> want to be an Ubuntu developer, I want to be an app author, but I want
> an on-ramp to harnessing the Ubuntu platform for my app.

The proposed process doesn't help that any.  It still needs to be packaged for 
a PPA and as far as has been revealed, the technical requirements are the 
same.

> So it seems many of us are in agreement of the problem to solve, and the
> process outlined here is one approach to solving this problem. Sure, it
> is not perfect, there are many implied assumptions (such as getting
> enough interest in staffing the Application Review Board! :-) ), and in
> the end it may not work at all, but I don't see how we can streamline a
> set of developer assessment processes that are designed for Operating
> System developers and make them suitable for folks who want to package
> one application.

Unless you change the requirements, which AFAICT hasn't happened, it takes 
equal packaging knowledge regardless of if the package is meant to be in 
Ubuntu or is aimed at this "not really Ubuntu, but in the software center".

> I also want to set expectations correctly about my involvement in this
> discussion: my goal was to look after and codify a new governance board
> surrounding application review for these apps. I am more than happy to
> discuss this process, which I put together, but I am not involved in the
> technical implementation of this feature -- I defer that to folks who
> are far more knowledgeable and suitable to have that discussion. :-)
> 
> I am keen for us to find a solution to the challenge of empowering
> application authors to get their applications in Ubuntu without being
> assessed with a set of processes designed for Operating System
> integrators. If this approach is unsuitable, can those of you who
> disagree with it suggest an alternative solution?

Already done.  

Scott K