Request For Candidates: Application Review Board

Rick Spencer rick.spencer at
Mon Aug 16 15:05:51 BST 2010

Good morning all, 

I'll try to respond to a lot of topics brought up in various threads in
my reply here.

On Sun, 2010-08-15 at 09:44 +0100, Matthew East wrote:
> According to Jono's original post, this process is not a Canonical
> formulated plan at all but has been developed with consultation with
> community contributors and the Technical Board. I can see that it has
> been discussed at this thread [1] and there also seems to have been a
> (private?) meeting about it [2] but it seems to me that the process
> that has been followed hasn't succeeded in getting feedback from
> people with an important opinion about it.
There were three rather large sessions at UDS, 2 to discuss
implementation, and 1 to discuss the process. Then all the work was done
in the normal transparent manner, in blueprints, specs, and the code was
published into Ubuntu as it was ready.

However, it is clear to me that subsequent discussion was inadequate.

At UDS there was a lot of feedback with a lot of different ideas and
strong opinions about how to implement this that were shared. In the
end, the people doing the actually work were, as should be, empowered to
make the decisions. It's not the case that there was a single unified
proposal put forward that everyone agreed to. Using backports, a new
pocket, PPAs, etc... were all proposed as possible implementations.
There are a lot of moving parts here, and getting something that
everyone could agree to for taking the first step was not easy
(software-center developers, launchpad developers, tech board, desktop
team, etc...)

I'd like to make clear that this initial implementation was put forward
as a "tech preview" fully with the expectation that we would revise the
implementation and process based on what actually happens after we try

> [1]
> [2]
> Is there anything that could have been done differently during the
> process which would have allowed these opinions to be taken into
> account before the process got to this stage? (I appreciate that Jono
> has said that the process is a "first cut" but it has at least got to
> the stage of asking for members of a board, and there appear to be
> objections of principle to the basic approach taken.)
I think the normal process for feature developed in the open was
followed here. However, it is clear to me in hindsight that the status
of the development should have been pushed, rather than pulled. I think
this list would have been a good place to do that pushing.

Please note:
1. I have read every bit of feedback on this thread, and I am not
resistant to it. Good points have been made regarding the
implementation, the proposed process, and the transparency with which
the feature was developed. All of this feedback will be acted upon.
2. The process (and therefore the implementation) is very much a
"technical preview" for Maverick, the Tech Board made very clear that
they expect the implementation and the process to be revised in Maverick
+ 1 based what we learn in Maverick.
3. Finding a way to allow application developers to deliver apps into
software-center for the current stable release in a reasonable time
frame is crucial to grow, or even maintain, the relevance of our
4. I haven't seen any objection to this goal (#3) in this thread, only
objections to the implementation, the proposed process, and the decision
making process.

I take full responsibility for the lack of sufficient communication
around the development of this feature area. I am also in a position to
ensure that we follow through on the original plans to revisit this for
NN, or OO after we have some experience with the Maverick

Cheers, Rick

More information about the ubuntu-devel mailing list