On 6/10/07, Daniel Stenberg <daniel_at_haxx.se> wrote:
>
> On Sun, 10 Jun 2007, Robert Iakobashvili wrote:
>
> > Let's say, somebody wants to emulate behavior of Firefox or MSIE, using
> > libcurl.
>

1) libcurl's use of connections is not designed after how browsers work or
> might work, since I don't know how they do it and I don't think libcurl
> needs
> to mimic browsers that closely.

Sure.

2) Browsers typcially does not use pipelining, or at least until recent
> years
> when Firefox added the ability to switch it on.

Experimental support for pipelining was in Mozilla already in 2002.

The world of pipelining is even larger.
Servers not necessarily getting requests from browsers. In many cases
there is an application/content acceleration device in the middle like
Radware, F5, etc made.
The devices are making multiplexing to decrease the number of connections
and, recently, also pipelining to be activated for those servers, that can
take
advantages of that.

3) The http spec recommends using no more than two connections to a single
> server, and I believe they intened the primary connection to be used to
> get
> the main html, and then you do a second connection (with pipelining) to
> get
> all the images etc that the main HTML refers to.
>

Sometimes, the real world is different and its practice finds its way to
RFCes

For libcurl, you would add an easy handle for each connection you want to
> make. libcurl's pipelining support doesn't currently allow you to do any
> fine-
> grained control on what gets pipelined or not, so if you enable it you'll
> get
> pipelining for every subsequent request sent to the same server.

Thank you for your advise.

-- 
Sincerely,
Robert Iakobashvili,
coroberti %x40 gmail %x2e com
...........................................................
http://curl-loader.sourceforge.net
A web testing and traffic generation tool.