phil wilson :: a geek commodity

3:15 PM

permanent link to post  

Wednesday, April 23, 2003

Sam Ruby’s last few posts about RSS past and future help to explain why it needs namespaces (or some other change) in the future, which helped clear my previous confusion. That makes sense, but for weblog aggregation? I'd still argue against a body tag, but what do I know?

Tim Bray wrote:

RSS is no longer a science experiment, it’s becoming an important part of the infrastructure, which means that a lot of programmmers are going to get the assignment of generating and parsing it, and they need better instructions.

In summary:

the two problems he complains about are that entity-encoded HTML is allowed and that relative URI references are not allowed.

I agree with Bill Kearney here, Before we bother with extending all this, let’s get the core stuff working first.. I think it would be fair to say that RSS is mainly used for website aggregation, and seeing as the apps that are out there don't even agree on how RSS should be parsed, and the specs are ‘conversational’ (I read that somewhere, probably in a comment on Sam Ruby's weblog, and liked it - conversational specs, heh), it might be nice to have some ratification and concreting of what we've got. Not to slow stuff down, but so we can draw a line and say this is the version websites should use and this is how it's supposed to work, and this is what you're doing wrong (of course anyone who wants to can use the latest in RSS feeds, and implement changes as they see fit - Sam does this well by providing a whole range of RSS versions for his site).

The part I disagree with most in all of this is that of allowing relative links. It’s probably the single most annoying practical problem around. There’s no consensus amongst RSS publishers or consumers on how relative links should be resolved. With respect to the server root? With respect to the <link> tag? With respect to the actual RSS feed?

A while ago, Scott Andrew wrote

Once again I broke my RSS feed with a relative URL. And once again I must ask: why can’t RSS readers resolve relative URLs like browsers do? I am after all required to include a base URL in the LINK element of my RSS feed. Surely I’m not the only one who thinks this is bogus. Route around it.

AFAIC links should be absolute or begin with a ‘/’ and be relative to the webserver root. How else do you specify where it should be? Relative links blow chunks.