On lör, 2007-12-08 at 18:32 +0100, Robert Siemer wrote: > 1) Where exactly did you find anything explicit for unknown methods? > Especially a forward-must... I can't find any, but have a strong feeling there should be a "a transparent proxy MUST forward.." and is also how I have understood the spirit of the RFC, just as there is such a rule for extension-headers in 7.1 entity headers. > 2) It would be new to me that the spec does force an implementation to > contain code for something that can perfectly be switched off. > Where does the spec require such a useless thing? More or less every SHOULD is in the above category (can be switched off). But it doesn't make them useless. Forwarding of extension methods is not useless. It's what allows the protocol to evolve with WebDAV, CalDAV etc, etc.. Regards Henrik
Received on Monday, 10 December 2007 05:07:28 UTC