- From: Balint Nagy Endre <bne@bne.ind.eunet.hu>
- Date: Tue, 16 Jan 1996 22:38:31 +0100 (MET)
- To: ddubois@rafiki.spyglass.com (Daniel DuBois)
- Cc: bne@bne.ind.eunet.hu, mogul@pa.dec.com, http-caching@pa.dec.com
> > >I think, we have no good reasons to prefer opaque validators over other > >validators. > >Current practice has two headers useable as validators Last-Modified > >and Content-Length (I forgot this when making the table.) If I say invalidators (especially when talking about Content-Length) then we are in sync? How to use those (in)validators, it's an other question. The situation is similar to content-negotiation: Some servers will support the use of last-modified as (in)validator, while others can support more. If a server itself doesn't support a validator, the user (data owner) can supply it by adding it as meta-information. We lose in that case the possibility of conditional GET's, but clients/caches can still use HEAD, and conditionally a GET, if (in)validators suggest that. Andrew. (Endre Balint Nagy) <bne@bne.ind.eunet.hu>
Received on Tuesday, 16 January 1996 22:05:38 UTC