Re: [syndication] Re: Finding Feeds

[Mark Nottingham <mnot@mnot.net>]

On Tue, Oct 02, 2001 at 09:56:09PM -0000, Bill Kearney wrote:
> > But this is not the point.  The point is that this would work in
> > a a distributed environment.  It would be nice to find feeds base
> > on the site similar to robots.txt.
> 
> Now this is an interesting idea.  Are you saying that when viewing a 
> web page that's also available 'somewhere' in syndicated format that 
> you want your browser to pick up on that and tell you?  Now, THAT 
> would be slick.  It would really 'complete the circuit' from fed 
> items to their website and back again.

That's what I'd like to see; see the "Linking to channel lists"
proposal in [1. I think the on-page link is an intermediate step,
until browsers support indicating a feed is there in the UI natively
(know any mozilla coders?).

I *think* we could probably just drop the 'channel list' file and
discover feeds based upon content-type, and dispatch them
appropriately.

In other words, when a browser sees

<LINK REL="rss-feed" TYPE="text/rss+xml"
 HREF="http://www.example.com/feed.rss";?

or a user follows a link that ends up at the same place, the browser
dispatches (based on content-type) the link to a piece of software,
which is either

a) a local aggregator 

b) a helper app that send the link to an online aggregator
(eventually, the browsers could be configured to do this themselves,
maybe sooner with a plugin)



> Or are you saying that the 'base' URL for a given site should have 
> something that helps the browser understand that a feed is available 
> for 'some' portion of the site content?

I'm less convinced that this is a good mechanism, but it may be
useful in some cases. Generally, the only reason you'd want to
disconnect the metadata from the data is when you need to know the
metadata *before* you request the data (like with robots.txt or
p3p.xml). 


1. http://groups.yahoo.com/group/syndication/message/1991


-- 
Mark Nottingham
http://www.mnot.net/