[RFC][PATCH]Speed improvement on the http transport: add keepalive support

[Goffredo Baroncelli]

On Friday 17 February 2006 03:33, you (John A Meinel) wrote:
> Goffredo Baroncelli wrote:
> > Hi all, 
[...]
> 
> If you are copying code from another project, it generally goes into
> 'bzrlib.util'.
Ok, this is not a problem

> And we used to actually use URLGrabber. I believe we stopped using it
> because it failed to work properly on Windows. And we preferred having
> only 1 code path. (I believe Aaron was the one who pushed for this).
> 
> I think Martin has a branch which would use pyCurl which supports
> keep-alive and some other good stuff (as long as you have curl installed).
> 
> So I'm not opposed to this, but I have the feeling someone else has
> already looked into it, and thought urlgrabber wasn't the way to go.

Thanks, I am started to work on bazaar after the deletion of urlgrabber; but I wasn't 
able to find the reason/bug why urlgrabber was removed: could you point me to the
message/reason ?



 
> John
> =:->
> 
> 
> 

-- 
gpg key@ keyserver.linux.it: Goffredo Baroncelli (ghigo) <kreijack at inwind.it>
Key fingerprint = CE3C 7E01 6782 30A3 5B87  87C0 BB86 505C 6B2A CFF9