Sean McGrath was recently looking for a tool that would “coalesce various standalone feeds into, aggregated feeds so that my client side aggregator sees it as a single feed” which seems like a perfectly reasonable request. And it’s one of many useful features that server-side aggregators could be doing. Others include normalising feeds so they have consistent levels of detail (e.g. with/without content), data cleansing (e.g. removing embedded markup), etc. There’s an essay I keep meaning to write here. It relates to separating out the data processing and presentation elements of RSS applications — everyone can benefit if serverside aggregators can pick up some of the heavy lifting in data processing and caching.
Anyway, whilst looking over the chumpologica just now, it occured to me that a nice feature in a tool like the one Sean is seeking would be to read a FOAF description and build an aggregated “My Friends” feed automatically. It’s only a matter of walking through the seeAlso’d FOAF files and picking out the foaf:weblog properties associated with each friend.
Allowing the level of harvesting to be configurable (e.g. include a friend-of-a-friend-of-a-friend’s weblog) would give an instant “my community” feed.
There are probably other variants of this, especially as the weblog property isn’t limited to people — it can be applied to Projects also.
3 thoughts on “Synthetic RSS Feeds and FOAF”
Comments are closed.
Me and a couple of people are developing personalblog. I had the idea of having a tile (we use struts/tiles) with the latest new of friends site.
http://sourceforge.net/projects/personalblog
[]s
Echo
That’s almost what the chumpologica code does – its ‘config file’ is the FOAF file at http://pants.heddley.com/logica/blogsources.rdf. The key thing is doesn’t do is follow seeAlsos to find weblog information, and I was thinking along those lines earlier today. It wouldn’t make sense to spider a bunch of FOAF files on each run of the RSS aggregator, but a second script could do a spidering run and produce a blogsources file like the one above.
I’ll be writing up the script that produces the chumpologica (and publishing the code) at http://www.hackdiary.com as soon as possible.
http://www.hackdiary.com/archives/000036.html – the writeup, as promised.