- From: David W. Morris <dwm@shell.portal.com>
- Date: Thu, 14 Mar 1996 00:23:46 -0800 (PST)
- To: HTTP Caching Subgroup <http-caching@pa.dec.com>
On Wed, 13 Mar 1996, Larry Masinter wrote: > I can't believe this scheme is practical. There are too many ways in > which your demographics could be unreliable by a factor of 'nnnn' > depending on how cache implementors interpret > cache-control: MAX-uses=NNNN > that this seems useless to me. Well, that argument could be advanced regarding most any solution or other aspect of the HTTP protocols. If the servers/clients don't implement carefully, interoperability will suffer. I think this is an interesting idea ... the question would be how many of the real customers Netscape, W3C, etc. are aware of deemanding some form of usage and/or demographic information would be satisfied with this as a widely available first step? I believe we're still waiting to hear from Ari. I note that MAX-uses=2 would cut the net load in half for this category of references with a pretty small loss of significance. Dave Morris
Received on Thursday, 14 March 1996 08:41:08 UTC