hc-httpclient-users mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Roland Weber <ossf...@dubioso.net>
Subject Re: https and aspx
Date Sat, 10 Nov 2007 18:18:56 GMT
ggcampbell wrote:
> After figuring out about the single header my self I found your reference to
> http.protocol.single-cookie-header
> But I found some weird sites that accept only single header cookies but send
> multi
> header cookies.

And why should they not? Cookie: and Set-Cookie: are two completely
different headers, which are interpreted and composed by different
code on either side of the connection.

> By turning on http.protocol.single-cookie-header the
> response headers
> were missinterpreted skipping the second and later cookies.

You must have gotten something wrong there. The single-cookie-header
in HttpClient is interpreted exclusively by the code that generates
the Cookie: header. The code that interprets the incoming cookies is
completely independent from that. Of course there are other possible
causes for not interpreting returned cookies correctly. The most
likely explanation is that you had the wrong cookie policy configured
for the respective site.
http://jakarta.apache.org/httpcomponents/httpclient-3.x/cookies.html
If none of the cookie policies work, please open bug reports in JIRA
and include the Set-Cookie: headers that were not interpreted correctly.

> httpclient.getParams().setParameter("http.protocol.expect-continue",true);
> Would it make any difference to make this the default behavior of
> HttpClient?

Of course it would make a difference - why should you ask for
that change if it wouldn't? And if it changes your code from
being broken to working correctly, it is sure to break other
people's code from working correctly to being broken. Besides,
development of 3.1 has ended. We will only fix severe bugs in
the old codebase. No enhancements, no cosmetics, and surely
no change in default behavior.
For the 4.0 codebase, all the examples I remember had that
parameter set to true. It comes with a performance penalty for
small payloads, since there is an additional round trip even
if the request and payload would fit into a single IP packet.

cheers,
  Roland


---------------------------------------------------------------------
To unsubscribe, e-mail: httpclient-user-unsubscribe@jakarta.apache.org
For additional commands, e-mail: httpclient-user-help@jakarta.apache.org


Mime
View raw message