W3C home > Mailing lists > Public > ietf-http-wg@w3.org > July to September 1997

Re: [Fwd: 301/302]

From: Dave Kristol <dmk@bell-labs.com>
Date: Tue, 29 Jul 1997 18:02:12 -0400
Message-Id: <33DE6864.4487EB71@bell-labs.com>
To: Josh Cohen <josh@netscape.com>
Cc: http-wg%cuckoo.hpl.hp.com@hplb.hpl.hp.com
X-Mailing-List: <http-wg@cuckoo.hpl.hp.com> archive/latest/3995
Josh Cohen wrote:
> Dave Kristol wrote:
> >
> > > Well, what about when foo.cgi is running on a 1.1 server, and
> > > doesnt give a content-length.
> > > Is the 1.1 server responsible to detect and chunk that ?
> >
> > An HTTP/1.1 server can either
> >         - send Connection: close, return the content without chunking, and
> >                 close the connection; or
> >         - chunk the content
> >
> Yes, but wasnt this the intent of chunking in the first place?
> (dyamic content)

We seem to be talking past each other, or something.  I don't understand
your remark.

Yes, the intent of chunking is dynamic content.  Well, actually, in my
mind the purpose of chunking is to provide a way to demarcate the end of
dynamic content so a server can keep its connection open.  So the
preferred behavior is that, if an HTTP/1.1 server supports CGIs, for
example, and the CGI does not provide a Content-Length, and the client
is also HTTP/1.1, the server can add a "Transfer-Encoding: chunked"
header to the response, chunk the output, and keep the connection open. 
But it's not required.

Dave Kristol
Received on Tuesday, 29 July 1997 15:06:52 UTC

This archive was generated by hypermail 2.4.0 : Thursday, 2 February 2023 18:43:03 UTC