But not thousands :-( May be it should be possible to configure "bad servers" which are "allowed" to respond without content-length.
> -----Ursprüngliche Nachricht----- > Von: Kalnichevski, Oleg [mailto:[EMAIL PROTECTED] > Gesendet: Freitag, 21. November 2003 16:14 > An: Commons HttpClient Project > Betreff: RE: Thounds of "Response content length is not known" in the > log file > > > Odi, > In this case I personally think that a warning message is > quite appropriate. > > Oleg > > -----Original Message----- > From: Ortwin Glück [mailto:[EMAIL PROTECTED] > Sent: Friday, November 21, 2003 16:11 > To: Commons HttpClient Project > Subject: Re: Thounds of "Response content length is not known" in the > log file > > > Kalnichevski, Oleg wrote: > > Odi, > > If I interpret the spec right, there's not much of a > difference. In HTTP/1.0 an entity enclosing response must > have 'Content-Length'. In HTTP/1.1 an entity enclosing > response must have 'Content-Length' OR 'Transfer-Encoding'. > It is not allowed to have neither. > > Yes, but the point is that some HTTP 1.0 severs send no > Content-Length > and just close the connection after the response. Even though > RFC 1945 > section 7.2 clearly requires a Content-Length. > > > --------------------------------------------------------------------- > To unsubscribe, e-mail: > [EMAIL PROTECTED] > For additional commands, e-mail: > [EMAIL PROTECTED] > > > --------------------------------------------------------------------- > To unsubscribe, e-mail: > [EMAIL PROTECTED] > For additional commands, e-mail: > [EMAIL PROTECTED] > --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
