How not to do RSS
Lifehacker posted a question about duplicate entries in an RSS feed so I looked into it, and the feed in question attaches a different tracking id to the item’s link and guid, each time the feed is fetched.
Brilliant! Google Reader assumes it’s a new entry and you get duplicates.
I’m using SharpReader and it seems to detect stuff like that (that is: it doesn’t trust the RSS data to be correct)… I’m not sure how I feel about that, but at least I’m not getting duplicates.
(via Scripting News)
July 12th, 2006 at 1:25 pm
[…] Patrick Fitzgerald sheds some light. […]
July 12th, 2006 at 6:27 pm
The brilliance is that it’s perfectly acceptable according to the RSS 2.0 specification. Duplicate items are unavoidable with (especially when aggregating from multiple sources) unless they are clearly identified. RSS 2.0 is totally sloppy on entry identification. But at least it’s simple, eh?
July 13th, 2006 at 9:37 am
> The brilliance is that it’s perfectly acceptable according to the RSS 2.0 specification.
According to the spec, the GUID is “a string that uniquely identifies the item”; however, these people are setting it to a random value each time the feed is retrieved. Since that’s not really a GUID, it should not be provided.
The GUID seems perfectly fine for entry identification, but since it can be misused (as in this case) some RSS aggregators (like SharpReader) choose to deal with it in other ways.