- From: Dan Connolly <connolly@w3.org>
- Date: Tue, 29 Jun 1999 22:35:50 -0500
- To: roconnor@uwaterloo.ca
- CC: www-talk@w3.org
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if the copy of your web page that I'm looking at or getting via my intercontinental caching proxy is 12 hours or so out of date, it's no crime against humanity.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