Re: [Monotone-devel] Re: comments on win32 patches

[Christof Petig]

Jon Bright schrieb:
> > Similarly, I'd be tempted to just have a spawn_and_wait interface, but
> > I suppose spawn_with_pipes needs something different... though we
> > should see exactly what's needed to support spawn_with_pipes(), hook
> > it up to netxx, etc.
>
>
> I definitely think spawn() and wait() should be different functions.  I
> can easily imagine hooks (note_commit, for example) wanting to spawn
> some external mail job or similar, and really not caring when it ends. I
> planned to look at Christof's stuff to get an idea of what's needed for
> spawn_with_pipes() - but it definitely doesn't want wait(), and I don't
> think it wants any links to netxx..

Actually it does. Netxx is the layer used by netsync and that is exactly
what the ssh branch covers.

For the cvssync branch I chose to implement first (using
pipe+socket+write+al.), port later, but I would have chosen Netxx
because I need a common abstraction for spawn_with_pipes (CVS_RSH) and
socket+connect (pserver) and was familiar with Netxx from the ssh branch.

So actually Netxx would be the layer to target (or any other socket+pipe
capable abstraction).

   Christof

signature.asc
Description: OpenPGP digital signature