On 6/21/07, Jeremy Whitlock <jcscoobyrs@gmail.com> wrote:
> Karl,
> Yeah. I had a discussion today on irc about this and while those
> are similar, I couldn't find an issue that was specific. 2619 points
> out a problem with how you background a process in your hook but the
> hook still waits for the backgrounded process to finish before
> returning back to the client, which really means you do not benefit
> from backgrounding processes in this context. Not what I'm talking
> about but it is probably related. 443 is more inline with what I'm
> proposing but it doesn't suggest that this functionality be available
> to *all* post-* hooks. It also seems to reference some issues that
> have been fixed, like the ability to return stderr for the pre-commit
> hook for example. Since all pre-* hooks have this functionality, I
> think that all post-* hooks would benefit from this functionality as
> well. If those issues (443, 2619) are a better place to discuss, I'll
> do it but this is more of a specific request I think. I've been wrong
> before though so feel free to point that out.
Hi Jeremy,
I think you're right and your request is more specific. However, since
you're also committed to create a patch for the issue, there's
actually no need to create one in the database yet. [And there won't
be at all, if you're prepared to see the submission process through.]
So, what you could do now is research what needs adjusting and where.
Write that down in a proposal for change. Then, when you explained all
aspects to the change (and nobody objects too much), you can start
coding.
Likely, the patch will see a number of revisions where we will ask you
to adjust specific aspects of the patch. This may be anything from
coding style to implementation details.
bye,
Erik.
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Thu Jun 21 08:46:33 2007