- From: Jeffrey Mogul <mogul@pa.dec.com>
- Date: Thu, 30 Nov 95 18:23:16 PST
- To: John Franks <john@math.nwu.edu>
- Cc: http-wg%cuckoo.hpl.hp.com@hplb.hpl.hp.com
I would also vote for an opaque cache validator. We would need an
"If-cache-valid: <validator> " and an "If-cache-stale: <validator>"
and we would need to specify their *semantics*. E.g. for a Range:
request with an If-cache-valid header send the range if valid, else
send entire document.
Can you explain why we need both
If-cache-valid: <validator>
and
If-cache-stale: <validator>
instead of simply
Cache-validator: <validator>
along with a set of rules that explain how it is supposed to
be interpreted?
E.g., for
GET
Range: 3-8
Cache-validator: XYZZY
I would expect the semantics to be
if the validator of the actual object is XYZZY then
return range 3-8, else return the whole thing
You might argue that one could use:
GET
Range: 3-8
If-cache-stale: XYZZY
but this seems to mean
if the validator of the actual object is NOT XYZZY then
return range 3-8, else return nothing
but that doesn't make a lot of sense to me, because I don't think
it is rational to obtain a range of bytes if your cached copy
of the entire document is known to be invalid.
-Jeff
Received on Thursday, 30 November 1995 18:31:27 UTC