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

Re: domain-name?

From: David W. Morris <dwm@shell.portal.com>
Date: Thu, 21 Sep 1995 12:13:11 -0700 (PDT)
To: http working group <http-wg%cuckoo.hpl.hp.com@hplb.hpl.hp.com>
Message-Id: <Pine.SUN.3.90.950921111836.4132B-100000@jobe.shell.portal.com>


On Thu, 21 Sep 1995, Lou Montulli wrote:

> I understand that Orig-URI is optional, and it doesn't suitably solve
> the very important problem of serving different content based
> on different domain names pointing at the same IP address. 
> 
> Let's forget about Orig-URI and concentrate on fixing this
> problem.  Sending Orig-Host solves the problem, and in my
> opinion is the best solution.

A possible compromise .. define Orig-URI such that in general use it
would be (almost) just the host. Something like:

   The Orig-URI header is used by the client to provide the server with
   those portions of the original URI removed when forumulating the 
   request URL. The client may either provide the complete original URI
   or it may include only the omitted prefix (protocol and host) and suffix 
   (e.g., fragment) by indicating an insertion point in the Orig-URI value
   for the request URL with a double per-cent (%%) (which is not a legal
   sequence in an encoded-URL). If the Orig-URI header value  doesn't 
   include a slash following the host portion and doesn't include %%,
   then %% should be implied on the end of the Orig-URI value.

For there to be a solution to Lou's content differentiation concern,
most clients must provide the original host fragment. I believe that
something similar to what I have proposed will resolve Lou's concern
while not exploding the HTTP request with special purpose headers as
the other issues Orig-URI is intended to resolve are addressed.

Dave Morris
Received on Thursday, 21 September 1995 12:17:56 EDT

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