RE: NEW ISSUE: Methods and Caching

On mån, 2008-11-17 at 10:00 -0600, Brian Smith wrote:

> Where is the discussion?

On this list somewhere..

> RFC 4918 says PROPFIND is cacheable. The only practical way PROPFIND can be
> cacheable is if the method is part of the key. Otherwise the cached PROPFIND
> responses would be returned by the cache for GET requests, which is clearly
> wrong. 

So nice of RFC4918, and works fine in context of a WabDAV client, but
not in general HTTP terms.

> You are proposing that 9(1) only GET and HEAD are cacheable, and in
> particular (2) POST and PROPFIND are never cacheable even with Cache-Control
> headers in the responses, and (3) that there will never be another cacheable
> method in HTTP. Off the top of my head, I can't think of any problems with
> that proposal as long as caches aren't caching PROPFIND already.

I don't propose dropping the cacheability properties of POST->GET, but
other than that yes.