On Sat, 12 Oct 2002, RiCH wrote:

> I tried some other URL's of much larger pages I could test easily, they did
> time out... but seems the issue is rather with the interpretation of a 1
> second timeout(below) not being 1 second. Maybe due to the sample time
> lengths the speed is calculated on being longer than a second?

The resolution used in the internal function for this check, is not very
exact. You should not count on it to be perfect. As a matter of fact, I
believe it might sometimes take almost a whole extra second to detect the
situation.

> Is there a possibility to make the timeout start at maybe 1/10 of a second?

Everything is possible, but it would take a change of libcurl for that to
happen. Personally, I'm not very keen on doing this change.

You have rather unusual request. I think you'd be better off by using libcurl
directly instead, then you could abort the transfer immediately without
waiting any extra time at all. There are mulitple mature and fine bindings,
even for your favourite script language, to choose from.

-- 
 Daniel Stenberg -- curl, cURL, Curl, CURL. Groks URLs.
-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf