Validate your feeds
by Volker Weber
It looks like the Domino folks are learning new tricks. RSS feeds this time. Gentlemen, please validate your feeds. Against both the validators from Dave Winer at UserLand.com as well as Mark Pilgrim and Sam Ruby at Archive.org.
This is not an option. It does not matter if your design breaks a few browsers, but a feed should never break an aggregator. It has to validate.
Comments
Being in the middle of a project to build a Domino-based aggregator, I agree -- but I also believe that practical aggregators have to be written to be as tolerant as possible of the variety of non-conformant feeds that our out there. My own feed fails the Archive.org validator -- which doesn't like the date tag (and I should surely fix that) -- but my aggregator code handles it (and many other non-conforming feeds) anyhow. I'm actually more concerned about the fact that even sytactically validated feeds have differing semantics which can be more difficult to deal with. For example, the best primary key to use for an item might be either the link or guid elements, or the rdf:about attribute, but you can't really count on any one of them being held constant for the same article over multiple feeds. There are quite a few other annoyances and complications that I've run into, and I've only been working on this part-time for about a week -- with about 75 feeds coming in from Domino and non-Domino sources. I've actually found that the feeds coming from some of the very prominent non-Domino bloggers are pretty useless, whereas most of the Domino-based feeds at least give me the information I need.
-rich
That is very true. There a lots of variations of the spec and different ways to express things. Most bloggers will simple go with the default templates that came with their software. Unfortunately the early MovableType templates were nonconforming (as is the current default template for the index pages).
Validating is the bare minimum requirement. As a user I would be really happy if the publishers would strip their tag soup from their feeds. I don't need (and don't want) their font tags.
Steve C and I have been working on fixing the problem with my feed. Fix should be in soon...
And this fix is just related to embedding of images by rich text - otherwise the RSS validates no problem.
Post a comment
Recent comments
Michael Sedlaczek
on Updates from vowe's Apple zoo at 19:34
Frank Koehntopp
on Updates from vowe's Apple zoo at 17:27
Tobias Mueller
on Updates from vowe's Apple zoo at 14:11
Lukas Praml
on Updates from vowe's Apple zoo at 14:01
Paul Farris
on Google syncs address book to BlackBerrys at 12:21
Alan Bell
on Amazing Head Tracking at 16:45
Oswald Prucker
on Google syncs address book to BlackBerrys at 14:36
Kai Schmalenbach
on Apple TV: First impressions at 11:13
Armin Roth
on Google syncs address book to BlackBerrys at 10:47
Gonzague Dambricourt
on Google syncs address book to BlackBerrys at 23:57
Nick Daisley
on Foldershare, R.I.P. at 20:20
peter mojica
on Notes or Exchange? Let's put an end to those creative statistics at 19:42
Armin Roth
on Weihnachtsmüll von Microsoft at 10:33
Michael Sedlaczek
on Weihnachtsmüll von Microsoft at 10:26
Julian Buss
on Use scripts to improve install process for Lotus Notes at 08:36
Philipp Sury
on Apple TV: First impressions at 07:04
Frank Paolino
on New poll: which BlackBerry could be your next? at 05:10
marco foellmer
on BlackBerry Curve 8900 at 00:20
Uwe Brahm
on Foldershare, R.I.P. at 19:03
Nick Shelness
on Foldershare, R.I.P. at 16:55
Nick Shelness
on Foldershare, R.I.P. at 16:43
Ralf M Petter
on Foldershare, R.I.P. at 15:32
Jan Tietze
on Foldershare, R.I.P. at 14:12
Ole Saalmann
on Foldershare, R.I.P. at 13:33
Stefan Weigand
on Doc Scrubber cleans your Word document at 13:21



