David Crossley wrote:
...
> Lets say that Apache XXX project takes on a new committer
> who is a loose cannon. I have actually seen this happen
> already.
This sucks.
...
> So in this hypothetical, that person now has access to
> Forrest SVN. They come here too and start making changes
> with no idea, nor care, about our desired direction.
> They just want to get the job done, their way.
>
> Yes, our PMC can always roll back changes and revoke
> access, but that causes community friction.
I don't think so, if we are fast enough.
If we are going to open to all, we *must* acknowledge that in this case
we *must* be swift and drastic, revoking access first and asking
questions later. After that, we must immediately contact the relevant
PMC and send a note to the board - and our part is done.
I have never seen this happen in all these years, mainly because loose
cannons usually never put themselves in a position to be easily
retaliated... which means that they will do things only in the projects
they can vote in, and in which they have already contributed stuff.
> Also with a small number of active people on our PMC, that oversight
> becomes difficult.
Once you see loads of commits coming in, you bet everyone will be
triggered right away.
> Anyway, i just want to ensure that we all, especially
> our new PMC members, understand the implications.
Yes, thanks for the clear and on-the-spot worst-case scenario.
--
Nicola Ken Barozzi [EMAIL PROTECTED]
- verba volant, scripta manent -
(discussions get forgotten, just code remains)
---------------------------------------------------------------------
