2008/4/14, Daniel Stenberg wrote:

> Sure it would work fine, but I think the progress callback is fine for this
> as long it isn't overused.

Even when the description is "Call progress function when
poll()/select() is interrupted" I feel that the real issue behind this
change is the recurrent "I want to abort libcurl current operation
with a signal to recover control immediately" ;-)

And that those that want it this way won't feel comfortable enough
knowing that there's a potential race condition in which the signal
won't have a nearly immediate effect. It seems that a one second delay
is too much :-)

As a matter of fact, Pierre's patch addresses the race condition
greatly minimizing the time window in which it can still happen, but
with the problem that it calls Curl_pgrsUpdate().

And since this is Pierre's second complete patch trying to address
this whole issue, I suspect that he is certainly decided to get this
straightened somehow.

Pierre, a working example for docs/examples showing its use would of
course be also a good idea, and could be used as proof of concept if
it is finally accepted.

-- 
-=[Yang]=-