Re: [syndication] Re: Finding Feeds

[Mark Nottingham <mnot@mnot.net>]

On Tue, Oct 02, 2001 at 03:27:11AM -0000, Bill Kearney wrote:
> 
> I wonder if anyone's done something like a
> soap://server:port/service protocol handler extension?  Being able
> to use a service that understands how to 'tell' your browser how to
> respond might be a start.

URI schemes aren't supposed to be used like that (BTW, SOAP doesn't
do it either; it's still http, or whatever else the underlying
protocol is). The normal way to 'tell' browsers this is Content Type;
unfortunately, we don't have very good penetration on this, probably
because it's so hard (for some people) to configure their server
properly.

There's been talk about registering a content type, but we'd need a
normative reference for the format (probably, RSS submitted to the
IETF as Informational). I wonder how hard it would be to get the
Apache people to bind .rss to application/x-rss (the 'x-' convention
is frowned upon, but until application/rss[+xml] is standardised,
there isn't much else, except application/vnd.rss, which I don't
think really applies).

In the long run, I'd hope that the browser vendors develop XML
dispatch configuration mechanisms, just like they do for content type
today; i.e., configure your browser to launch example.app when it
sees an xml document with namespace 'http://www.example.com/'.


> One large hurdle is independence and it's not really surmountable. 
> Most sites and services actually want to remain independent.  We've
> all seen what happens when a service implodes (dejanews?).  That
> and fundamentally it's easier to program your local service if it
> only interacts with things under your direct control.

Right. Doing without having some centralised registry of feeds or
redirection service is core, to me.

Cheers,



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