I am looking into providing a demonstration of the Atom binding ETags
and Last-Modified support (see Jira TUSCANY-2537) using the
feed-aggregator sample. I intend to show the savings in network
bandwidth and business data over-the-wire as one does multiple get feed
requests with several lengthy feeds.
One area in particular is interesting is generating the feed id and
updated dates for the aggregated feed. (These are required elements of a
feed as required by the Atom publish protocol). I intend to come up with
the updated field of the feed by looking at the updated elements of each
entry and using the most recent update. I intend to come up with the id
field of the feed by taking a hash code of the feed and its contents.
This bit of calculation might take time to generate the feed id and
updated fields, but I feel it is needed for Web 2.0 style caching support.
Does anyone have alternate methods or pointers to other areas where
there are unique IDs and update fields for collections either inside or
outside of Tuscany? It would be useful to see how others are handling
these types of caching tasks.
--
Thanks, Dan Becker
- Atom binding caching demonstration for feed-aggrega... Dan Becker
-
