Calculation of age headers

I'm hoping we can resolve this soon, by widening the discussion
from "Roy vs. Jeff" to a larger group. It would be very useful
for HTTP/1.1 implementors to be explicit about their plans and
expectations. Is there some "conservative in what you send
and liberal in what you accept (or expect)" possible compromise

Comment from Ari:

> ---
> I much prefer Fielding's draft where all proxies that generate the
> response from their cache must add an Age: header.
> I don't want to require *all* proxies, including non-caching ones, to
> add the Age: header, which is what Mogul's draft suggests.  Since it's
> just tunneling it doesn't really add any information that the upstream
> proxy wouldn't be able to derive from its network delay.  Furthermore,
> this wouldn't affect just non-caching proxies, but also requests that
> we'd like to tunnel without caching in a caching proxy.

Received on Saturday, 27 September 1997 23:00:34 UTC