in reply to Re^2: RSS feeds to most of
in thread RSS feeds to most of

Good stuff - well done. I have added this as a live bookmark straight into Firefox.

You mentioned that your process used a daemon that checks for new nodes. This means that firstly, you need to run a daemon and secondly, you are interrogating PM on a regular basis.

You could simplify the model by caching the RSS for a particular page and interrogating the cache each time you wanted to serve a page. A cached page could time out after a short period of time (e.g. 10 mins). A cache miss (or timed out page) would initiate a request to the monastery. The result would be cached for next time. This means that when nobody was using the feed, PM wouldn't be hit.

Caching can be implemented using a simple file cache with timestamp checking or something more involved using a database. Either way, you periodically need to clean the cache of expired documents. You would also want to guard against an attack where a malicious user tried to access every node as a feed and therefore used up lot's of cache space.

This article may be of interest with respect to the database solution.

  • Comment on Re^3: RSS feeds to most of

Replies are listed 'Best First'.
Re^4: RSS feeds to most of
by EvdB (Deacon) on Mar 03, 2005 at 16:57 UTC

    I've added a page that describes (quickly) what is going on here. As you can see I am caching stuff already. It turns out that it is not possible to get the latest nodes or perlmonks on demand as it is not possible to know which nodes are needed for things like generating RSS feeds for threads.

    Thanks for your comments, I'm glad that you like it.

    --tidiness is the memory loss of environmental mnemonics