Re: Pre-fetching some URLs

#define quoting(Duane Wessels)
// purcell@rmcs.cranfield.ac.uk writes:
//
// >As a suggestion, it may be a good idea to have ttl_pattern (or what it is
// >called once 1.1 moves away from ttl) uncommented by default as this will be
// >what will be initially installed and operated. Then if objects are
// >remaining stale in the cache for too long ttl_pattern can be tweaked after
// >installation.
//
// I have never gotten mail saying "Wow, you have choosen perfect defaults!"
// We only get feedback when someone thinks the defaults are wrong.
//
// If this is a big problem, then we should form some sort of group (with
// votes etc.) to select appropriate default values.
//
// My experience is that people (especially users) would rather have
// fresh objects from half-full caches than stale objects from full ones.

I think that why the IMS was created for... You could specify a small
time to cache, and when expired use the IMS just to check if it's still
valid. Of course, this will only work if the cache cleanup is done
only on requests or when there's no more swap space. No timeout for
URL expiration should be used. Of course, this is JMHO, and I don't
know if squid works this way.

And, after all, if you you have a big timer cache configuration, and think
your page is stale, simply click reload and the <No Cache> request pragma
will force a reload. It works for me, and I'm very happy. (If this
matter, I use the NetScape browser)

//
// Duane W.
//
//

                                        Jonny

--
Joao Carlos Mendes Luis			jonny@gta.ufrj.br
+55 21 290-4698 ( Job )			jonny@cisi.coppe.ufrj.br
Network Manager				UFRJ/COPPE/CISI
Universidade Federal do Rio de Janeiro