- From: Alex Rousskov <rousskov@measurement-factory.com>
- Date: Tue, 25 Nov 2003 10:26:45 -0700 (MST)
- To: Jeffrey Mogul <Jeff.Mogul@hp.com>
- Cc: Julian Reschke <julian.reschke@gmx.de>, ietf-http-wg@w3.org, w3c-dist-auth@w3c.org
Given Jeff's comments, would it make any sense for those who need a working OPTIONS to specify their own extension method (e.g., FEATURES) and publish a small RFC instead of patching RFC 2616? Draft-ietf-http-options could be a starting point. One problem would be with proxies that do not pass extension methods on, unfortunately. But some do not pass OPTIONS as well, so it is not such a "new" limitation. Adding some extension header to OPTIONS, to signal new well-defined functionality would be a similar approach that does not require an extension method. Alex. On Tue, 25 Nov 2003, Jeffrey Mogul wrote: > I think there is no doubt that many or most of the authors think > that OPTIONS is basically broken in RFC2616, so a well-designed > rewording is probably welcome. However, given that (1) we are not > likely to change the version number at this stage, and (2) without > changing the version number we cannot really expect to place new > requirements on servers, any change that might be proposed is > probably more likely to be of the form "clients cannot count on > OPTIONS behavior, but here's some advice to server implementors." > > -Jeff > > P.S.: for what it is worth (not much at this point), Scott > himself was a co-author of draft-ietf-http-options-02.txt, > "Specification of HTTP/1.1 OPTIONS messages", back in 1997. My > recollection is that the Working Group rejected this approach and > so the I-D died, but in my opinion it was the best design that was > available. (My opinion is biased, for reasons that might be > obvious to anyone who reads this draft, available via Google and > other fine search engines.) >
Received on Tuesday, 25 November 2003 12:27:54 UTC