On 2 Sep 2014, at 09:17, Kevin Smith <[email protected]> wrote:

> There's an assumption running through a lot of posts in this thread
> that moving to a github-like pull request model would be a good thing.
> Our situation is somewhat different to the typical OSS project hosting
> on github. The primary responsibility for XEPs lies with the authors
> (Council might suggest new authors for XEPs where the original authors
> are AWOL, but a) that's rare and b) we don't have a procedure for that
> (about which I'm uncomfortable)), so it's not a case of submitting
> pull requests to a central body and waiting for them to get merged,
> those changes should come from the authors. This complicates 'throw it
> on github, we'll get exposure and patches and the world will be full
> of unicorns and fairies' significantly.

I can imagine a world of fairies and unicorns where (for example) we have an 
(openid enabled) gitlab, sufficiently automated that we have a repo per xep 
which is writable by the authors, but anyone can raise issues and pull requests 
against the xep. New issues and pull request notifications are sent to all the 
xep authors and copied to the standards list too.

This should be eminently automatable (is that a word?) and fulfils a couple of 
requirements - firstly it fits the current process of xep authors essentially 
being the single point of responsibility for changes to the xep, but also 
allows an issue tracker per xep (which could be useful in itself and makes it 
more obvious which issue applies to which xep) and provides an easy route for 
anyone to raise issues or corrections.

—
Ash

Attachment: signature.asc
Description: Message signed with OpenPGP using GPGMail

Reply via email to