W3C home > Mailing lists > Public > ietf-http-wg-old@w3.org > September to December 1995

Re: domain-name?

From: Lou Montulli <montulli@mozilla.com>
Date: Thu, 21 Sep 1995 10:42:04 -0700
Message-Id: <3061A3EC.41C6@mozilla.com>
To: Roy Fielding <fielding@beach.w3.org>
Cc: Lou Montulli <montulli@mozilla.com>, Chris Schefler <css@webcom.com>, http-wg%cuckoo.hpl.hp.com@hplb.hpl.hp.com
Roy Fielding wrote:
> 
> >> Questions:  Is there anything in the 1.1 spec for passing the full
> >> URL, or, alternatively, passing the domain name as a header?
> >> Or am I correct in my reading of the spec that passing the full
> >> URL, including domain name, is legal in 1.0? (e.g., GET
> >> http://www.domain.com/dir/welcome.html HTTP/1.0)?
> >>
> >There is an ORIG-URI header that can contain the fully specified
> >URL.
> 
> Yep, that's right.
> 
> > I find this header to be very wasteful though.  I would
> >prefer to see an ORIG-HOST header that would only specify the
> >hostname so that it isn't necessary to transmit the path
> >part of a URL twice.
> 
> What then would you include for the path URI?
> Orig-URI will be useful when URNs show up (if URNs show up).
> 
> All requests will include the full URI in HTTP/2.0.
> Until that time, Orig-URI is sufficient.
> 

It is not at all sufficient.  It requires doubling the amount
of request data sent for long paths.  If I submit a form with
30 K of data with a GET method, I certainly don't want to
send 60K of data.  If we send the missing parts of the
URI as headers they can be reassembled at the host end.
There is no need for something as wasteful as Orig-URI.

:lou
-- 
Lou Montulli                 http://www.mcom.com/people/montulli/
       Netscape Communications Corp.
Received on Thursday, 21 September 1995 10:48:56 EDT

This archive was generated by hypermail pre-2.1.9 : Wednesday, 24 September 2003 06:31:32 EDT