> From: Stuart Roebuck [mailto:[EMAIL PROTECTED]] > > > On Wednesday, May 22, 2002, at 01:37 PM, Berin Loritsch wrote: > > >> From: Stuart Roebuck [mailto:[EMAIL PROTECTED]] > >> > >> My only concern here is that publications don't need to be checked > >> synchronously for general use, but they may need to be checked > >> synchronously as part of implementing a content management > system for > >> people who are accessing the content immediately after a change. > > > > > > If your CMS manually invalidated the entry for that resource when a > > change is checked in, then your problem is solved. The > asynchronous > > checking is more advantageous for filesystem resources and > such. The > > fact that the CMS checks things in using another thread, (as all > > requests are in separate threads), gives the effect of asynchronous > > management. > > Do you mean that the CMS manually removes the cache entries? If it > invalidated the entries this would still hit against the asynchronous > validation issue as it would still have to determine that the entries > were invalid, or am I misunderstanding things?
With cache systems, invalidating the entry means that the cache will remove it, and serve the new updated information in its stead. However this is in a write/read-back style cache where the cache lives in between the component and the resource. --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, email: [EMAIL PROTECTED]
