Archive of UserLand's first discussion group, started October 5, 1998.
An RSS Categorization Proposal
I wanted to start a new thread since the other one was getting scary.
I also wanted to comment on the excellent discussion, which is not only an issue for RSS, but is a big issue on the Web in general.
First, I think it's safe to say that there shouldn't be a definitive
categorization scheme for RSS or the Web. I've used systems like the Standard
Industrial Classification (SIC) codes and the Dewey decimal system, among
others, to categorize data on the Web. While these systems are useful, they
do not provide the granularity and detail necessary for classifying data.
For example, while a fruit stand and an apple orchard both fit under
the agriculture category, the information that's valuable to a fruit
stand owner will be different than that of an apple grower. Categorization
is best left to the experts of that domain.
I think John Udell had it right when he suggested a hybrid approach.
One nice thing about XML is that you can import properties from many other
namespaces. So, we could effectively include as many categorization
schemes as we like. The tags might look something like:
005.7
News/By_Subject/Internet
Xml-rpc to the future?
Markets & Investing/Personal Finance
This way, we can allow groups or individuals to use their own
categorization schemes. As groups release new schemes, you
can decide whether to update your software to recognize the
new namespace or simply ignore them. Also, content providers
would have multiple schemes to pick from, so scripting.com
might include info for dewey, dmoz, and its own scheme.
Finally, I would recommend that you read the XMLNews Meta
tutorial at
http://www.xmlnews.org/docs/meta-tutorial.html, which has several examples of importing
namespaces. It's also a much richer format that RSS IMHO.
Jonathan.
There are responses to this message:
This page was archived on 6/13/2001; 4:52:34 PM.
© Copyright 1998-2001 UserLand Software, Inc.