Open Stack

On 3/18/14, 8:18 AM, "Sean Dague" <sean at dague.net> wrote:

>
>As long as this is going to be evaluated strongly at the end of the
>process (including for the power user), I'm fine with that, and will
>drop it at this point.
>
>	-Sean

Our intent is to not put out something that will not fit the needs of the
team as a whole. It is true that I have not been that involved in commit
reviewing, but one of the developers working on this project with me is in
the higher up in the review count. He is currently sitting at 660+ reviews
in the Icehouse cycle so he has a lot of experience with Gerrit.

I like the idea of taking a step back and starting with the basis of a
well defined UI around Gerrit as a first step, but I think this should be
a small portion of the process as working towards a backend that can help
our pipeline will be a great thing.

What I am thinking might be the best approach is a two pronged attack to
start: 

1. Start on an AngularJS wrapper to Gerrit that is purely based around a
REST API.
2. Begin with a common core that can slowly be extended to provide
replacement to the Gerrit REST api along with filling in where we have
issues with our current pipeline.

-Ph