Re: [syndication] RSS Auto-Discovery 2.0

[Joe Gregorio <joe@bitworking.org>]

Jeremy Zawodny wrote:
> I'd like to suggest that we start thinking about the next level of RSS
> Auto-Discovery.  I've written up some initial thoughts and rationale
> here:
>
>   http://jeremy.zawodny.com/blog/archives/000967.html
>
> Do others think this would be helpful and relatively easy to spec out
> and implement.  Are there other (better) ideas floating around?  I've
> not seen much discussion on the topic...
>
> Jeremy

Interesting idea. I would rather not see the use of OPML, but
rather would like to see multiple link tags on the same
page, by loosening the restriction on title. We already
have the type="" attribute to indicate that the link
is for an RSS feed, so you could provide multiple
link tags on a page and set an appropriate title for
each one. When subscribing the user can be prompted
with those title for which feed they want to subscribe to.
For example, a top story that fell into both the
science and technology categories might have the following
link tags:

<link rel="alternate" type="application/rss+xml" title="Technology" 
href="http://rss.news.yahoo.com/rss/tech";>
<link rel="alternate" type="application/rss+xml" title="Science" 
href="http://rss.news.yahoo.com/rss/science";>
<link rel="alternate" type="application/rss+xml" title="Top Stories" 
href="http://rss.news.yahoo.com/rss/topstories";>

This has the benefits of:
1. Backward compatible with old aggregators,
   which will most likely just choose the first feed.
2. Does not introduce a new file format.
3. Does not require a second round trip to retrieve the OPML file.

	Thanks,
	-joe

--
http://BitWorking.org
http://WellFormedWeb.org