What’s the purpose of Atom?

I’ve started work on my RSS Manager again, and am now cutting code, which led me to put a _lot_ more thought into what data I was going to receive. Originally I was thinking only in terms of processing weblog RSS feeds, now I’m realising that a more general data aggregator would be a lot more useful.

And deciding on the data format to model the database tables on is proving tricky. I am starting to think that following RSS is going to be too limiting; to follow “RDF (there must be a better tutorial)”:http://www.xulplanet.com/tutorials/mozsdk/rdfstart.php too difficult as it’s extensible (someone *please* correct me if i’m wrong) unless I can create virtual tables for each feed, and map all the data to the display model.

Then I found “Atom”:http://www.intertwingly.net/wiki/pie/FrontPage. Is this the answer? Not sure, is my current reply. Here’s a few questions I haven’t found the answer to yet (anyone else find there wiki hard to find _information_ in?).

* Is Atom a generalised format for marking up any content, or specialised just for weblog content? The impression I get is that is has been designed with blogs in mind.
* Is it a subset of RDF or something separate?
* What do the developers see it being used for?

Data format is an important topic, and one which looks like it’ll hold up my development as I don’t want to continue until I’ve got it right. Below is shown the current database structure, based on that used by “Feed on Feeds”:http://minutillo.com/steve/feedonfeeds/. More fields will certainly appear; I’ve got a few lined up for the @Feed@ table already. Feed items will obviusly have to wait until the data structure is worked out.

!http://peter.mapledesign.co.uk/images/rss_manager_db_model.png!

4 comments ↓

#1 Sam Ruby on 03.08.04 at 1:57 pm

Since you are interested in feed formats from a data modelling perspective, the following two presentations may be of help:

http://www.intertwingly.net/slides/2003/seybold/
http://intertwingly.net/slides/2003/xmlconf/

#2 Peter Bowyer on 03.08.04 at 10:23 pm

Thanks Sam, the second link I found useful to explore Atom further.

Can you explain though what the difference is between RDF and Atom? The former was shouted around a few years ago as the ‘best format’, and seems to still be very widely researched into at University; but I cannot find any good online documentation into it.

#3 Sam Ruby on 03.10.04 at 4:17 pm

In essense, the difference between a database and a schema.

RDF is a mechanism for unifying everything into triples. The middle “column” is the predicate, and by defining new predicates, one creates a vocabulary.

RSS 1.0 is an RDF vocabulary expressed using the RDF/XML serialization. Neither RSS 2.0 nor Atom are RDF. They are simply “just XML”.

#4 Peter Bowyer on 03.10.04 at 8:47 pm

Again, thanks Sam for the explanation. I’m doing some reading into what ‘triples’ are (mostly stuff I’m finding on the “Redland”:http://www.redland.opensource.ac.uk/ site) and then hopefully be in a more informed position to decide what to use. Looks like it will be a RSS reader alone to start with; with potential to expand (should I be lucky enough to be allowed on a RDF-related course).

Leave a Comment