W3C home > Mailing lists > Public > ietf-http-wg-old@w3.org > May to August 1996

RE: Sections 3.3.1 and 5.1

From: Paul Leach <paulle@microsoft.com>
Date: Fri, 31 May 1996 14:32:16 -0700
Message-Id: <c=US%a=_%p=msft%l=RED-77-MSG-960531213216Z-5199@abash1.microsoft.com>
To: "'ben@algroup.co.uk'" <ben@algroup.co.uk>
Cc: "'http-wg%cuckoo.hpl.hp.com@hplb.hpl.hp.com'" <http-wg%cuckoo.hpl.hp.com@hplb.hpl.hp.com>


>----------
>From: 	Ben Laurie[SMTP:ben@gonzo.ben.algroup.co.uk]
>Subject: 	Re: Sections 3.3.1 and 5.1
>
>
>Proxies are permitted to rewrite internally. I understand the reasoning
>but
>it seems to me that this should be a restrictions on URLs (that is,
>that they
>are always in their canonical form) rather than on proxies.

That "permission" is actually ooutside the scope of the spec -- any HTTP
app is permitted to do anything it wants internally as long as it
conforms to the requirements of the protocol. That would perhaps be
clearer if the sentence

Proxies MAY transform the Request-URI for internal processing purposes,
but MUST NOT send such a transformed Request-URI  in forwarded requests.


were moved into the Note: that immediately followed (and the MAY and
MUST NOT changed to lower case, since MAY doesn't apply to the protocol
and the MUST NOT is redundant with the proscription on rewriting
expressed earlier in that section).

You suggestion about requiring canonical form has the right intent, but
since many existing clients don't send in canonical form, and the
purpose we were seeking (authentication) only requires that the URLs not
be modified, it seemed like an unnecesary burden to require canonical
form.

Paul
Received on Friday, 31 May 1996 15:10:16 EDT

This archive was generated by hypermail pre-2.1.9 : Wednesday, 24 September 2003 06:32:01 EDT