Re: mechanisms for reviews needed

[Derek Price]

Jacob wrote:

> You may claim none of this belong in a VCS and I agree in principle.
> Problem is that the system that should take care of this *outside*
> the VCS must be so tightly integrated with the VCS that it would be
> simpler just to define it as part of it. 


I believe integrating this exactly as you specified would slow your
developers down considerably.  If you really want something
approximating this behavior, I suggest that you use something like
cvs_acls to restrict commits to "clean" branches, but if you can't trust
your developers to test their code before they commit it to the correct
shared branch or to back out the occasional mistake, you have a
different problem.

If you are trying to integrate a large number of unknown developers, as
often happens on open source projects, perhaps what you are looking for
is an issue tracking system like bugzilla that allows patches to be
attached to issues and assigned to developers who are allowed to commit,
who you trust to review, test, and commit acceptable work.

Regards,

Derek

Get CVS Support at <http://ximbiot.com>!