Best practices for developing a feed aggregator / proxy?

Question

I'm looking into writing a proxy aggregator for feeds, where a handful of users specify a feed URL and some set of conditions, and the system outputs a continually-updated RSS/Atom feed of entries that match those conditions.

Are there best practices for feed aggregators? (Or filtering feed proxies?)

For example:

• Are there certain feed elements that should or shouldn't be modified be proxies?
• How should a feed proxy/parser indicate that it's not passing along a pristine copy of the original feed?
• Does it make sense to delegate the work of downloading/updating to a third party aggregator platform, e.g. the Google Feed API? I presume that'll save a lot of work, vs. having to do updates, 301 handling, etc. by myself.

Thanks for your help.

Answer 1

Don't get bought by Ask.com

Answer 2

Do not query any feed more frequently than 30 minutes. Use caching.

Answer 3

You could use also Yahoo Pipes, I guess... Or this one: planetplanet.org