On Mon, Jan 25, 2010 at 02:05:41PM -0800, Brock Pytlik wrote:
> To help me understand what was going on, I created the attached
> python program. Here's the output when I run it:
> running
> k:4
> Got to finally, i:1
> calling but not iterating
> after calling but not iterating
> k2:3
> Got to finally, i:0

You forgot to attach the program. :P

> b) The question of what to do if the caller doesn't run to
> completion is interesting, and I don't have a good answer.

Yeah, that's part of my concern. This entire idiom doesn't lend itself
well to locking.

> Your wrapper has the same semantics (I think) as putting the
> try/finally inside remote_search. What you have shouldn't cause the
> results to be held in memory, and it has the same problem with
> callers that don't run till completion.

Indeed.  I proposed a half baked solution.  I'm not quite sure how to
write this is a way that is safe for locking, and works incrementally.

-j
_______________________________________________
pkg-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/pkg-discuss

Reply via email to