Re: forced caching of volatile data

[Geoff Howard]

Hunsberger, Peter wrote:
Geoff Howard <[EMAIL PROTECTED]> writes:

Again, no change to the impact on the cache.  Just less Cocoon code:
everything is considered for caching, no key or no validity, no caching.
Same as today; but currently it's possible not to be even asked for a
key or validity. Going forward you'd always be asked (or the abstract
code would always be asked) but you could still decline to be cached.
By itself, the component still doesn't cache even when in a caching
pipeline.  Only when the expiry is sent is the pipeline 
bludgeoned into 
the cache.  Would this still be considered a violation of 
the contract?

Another problem I see with this is it locks into a time-based 
(expires) 
cache validity.  
Umm why?  You can still invalidate the validity in any manner you see
appropriate, or for that matter, you can still manage the cache of
actual data in any way you see fit...
I may have been jumping to conclusions, but my assumption based on the 
overall discussion was that to make forced caching easier that most of 
the Impl components would have a validity provided.  For instance the 
discussion on the added complexity of configuring my solution.  I could 
be misunderstanding here, though?

Good point on the sensitive data issue - I assume your work has similar 
requirements? (from what I remember)

Geoff