Re: RSS 0.92 Spec
--- In syndication@y..., "Dave Winer" <dave@u...> wrote:
> In some kind of ideal world every weblog item would contain a
> title, a link and a description, but you rarely see that, because
> HTML is the dominant format and it allows every word to be a link,
> and this is something most weblog authors take advantage of.
I agree. As you may or may not know, I run a news portal site at
http://www.newsisfree.com/ so I spend most of my evenings (this is
not my main job) writing software which parses syndicated news feeds
and renders them.
> My first attempt to deal with this, in an imperfect way, was in
> <scriptingNews> format in late 1997 and still in use today. Here's
> an example of my own weblog as viewed through this format:
yeah, I read that every day:
> This approach is imperfect, because it sometimes matches the wrong
> text with the link when it's put back together on the aggregator.
> But you do get the text separated from the links, but what do you
> gain from that? Hard to tell.
> A simple parser on the aggregator side could do the same thing
> with the source text.
> RSS 0.90 and 0.91 happened to quickly to even begin to address
> this concern and weblogs were still pretty new then.
yes, I guess this all makes sense. "usual" news feeds will have one
link and one text item. Rich ones will have several links. Maybe it
would be a good idea to limit the use of HTML to the <a> tag and
discourage the use of HTML formating in the RSS feeds...