Re: Static? Dynamic?

Russell Steven Shawn O'Connor wrote:
> > The problem is: so few pages declare their TTL that caches
> > have to assumes something like 24hours, and so providers
> > of dynamic services have to commit all sorts of hackery
> > to work around those caches. Sigh.
> 
> Of course declaring TTL is hard to impossible to do.

Knowing when your page will change may be impossible;
declaring a TTL *anyway* is not; witness the fact
that every DNS record declares its TTL.

i.e. don't look at TTL as "a time until which this page
is guaranteed not to change" but rather "a time until
which it's pretty much OK if you're still looking at
this version."

>  For example, I don't
> know when I'll next update my web page.

And yet, if the copy of your web page that I'm looking
at or getting via my intercontinental caching proxy
is 12 hours or so out of date, it's no crime against
humanity.

>  Whenever I feel like it.  Of
> course it's not that important if my page is a little stale in the cache,

Exactly.

> but it could be important.


As a community, I feel we have the following choice:
	-- provide TTL metadata in those cases where
		we don't want clients (incl. proxy caches)
		to assume a default TTL around a day or a week
		or whatever, so that folks can implement caching per
		the specs and get reasonable application behaviour
	or

	-- live with the cache-busting techniques that providers
		are forced into due to the current lack of discipline.

-- 
Dan Connolly, W3C
http://www.w3.org/People/Connolly/
tel:+1-512-310-2971 (office, mobile)
mailto:connolly.pager@w3.org (put your tel# in the Subject:)

Received on Tuesday, 29 June 1999 23:35:52 UTC