- From: Roy T. Fielding <fielding@avron.ICS.UCI.EDU>
- Date: Wed, 30 Nov 1994 05:25:22 -0800
- To: Marc VanHeyningen <mvanheyn@cs.indiana.edu>
- Cc: http-wg%cuckoo.hpl.hp.com@hplb.hpl.hp.com
Marc VanHeyningen writes: > Generally, I think the draft of Nov 28 is very good. Thanks. > Rather egregiously missing is a reference to transmitting network > objects in canonical form. Section 3.2 should mention this; a > reference to the canonical encoding model in Appendix G of RFC 1521 > (specifically step 2) probably should suffice. The only place this is > hinted at is in the tolerance section of the appendices on tolerance > of broken implementations, but the spec should explicitly say what the > proper behavior is, just in case any servers every actually do that. :-) The specified behavior will be "no canonical encoding of the object-body is required before network transfer via HTTP, though gateways may need to perform such canonical encoding before forwarding a message via a different protocol. However, servers may wish to perform such encoding (i.e. to compensate for unusual document structures), and may do so at their discretion." > As near as I can tell, the spec constrains all header values to be > US-ASCII, meaning nothing that is not US-ASCII may be contained in > them. We might consider permitting non-US-ASCII information in at > least some headers, probably using RFC 1522's model. I'd rather not. If there is a perceived need for non-US-ASCII information in header field value text and comments (I don't see any), then I think they should only be encoded by gateways during export. > In section 7.5, I don't understand the BNF for the CTE header. CTE's > don't have subtypes or parameters. Just trying to improve the world we live in. ;-) Okay, okay, I'll fix it. > Regarding busy server errors, a "Retry-After:" field might be > reasonable, but I would prefer to just make it an HTTP-date rather > than inventing something new for clients to have to parse. If we were > going to use relative dates, there are plenty of other places (like > Expires:) where they make as much sense. A pointer to an alternative > address also seems like a sensible way to handle timeouts. Actually, I was going to propose adding an optional delta-seconds for Expires as well, but not until HTTP/1.1. ......Roy Fielding ICS Grad Student, University of California, Irvine USA <fielding@ics.uci.edu> <URL:http://www.ics.uci.edu/dir/grad/Software/fielding>
Received on Wednesday, 30 November 1994 06:05:38 UTC