On Wed, 21 Oct 1998, Nottingham, Mark (Australia) wrote:
> As far as I can tell, an ETag request header is invalid; 14.9 states only
> that it is a response header.
ETag is an entity header. As such, it can be used in a request.
> AFAIK, you won't see much If-None-Match traffic at all; no clients seem to
> support Etags (I may be wrong).
Too bad if so -- less incentive for us to add ETag support.
> However, the specification doesn't preclude
> (and actually encourages, in my understanding) caches using ETags as
> validators when they're available.
True.
> As TODO notes, it will mean storing more metainformation.
That remains to be seen. Clearly some ETag applications will require more
metadata, but maybe not all. We can make such suport #conditional and/or
optimize it somehow...
Alex.
Received on Wed Oct 21 1998 - 00:46:34 MDT
This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 16:42:37 MST