Re: [syndication] friendlier feeds

["S. Mike Dierken" <mdierken@hotmail.com>]

The stuff I did was for a non-file based storage system, like running sql
against a db. I didn't think file extensions were the way to go for that
kind of dynamic system. That way the URI didn't change. (aside from query
terms & I was always fuzzy on whether those were considered 'identifiers'
for the resource.

But your example is a better (easier) way to do it when you've got files or
things that look like files.

----- Original Message -----
From: "Mark Nottingham" <mnot@mnot.net>
To: <syndication@yahoogroups.com>
Sent: Tuesday, January 01, 2002 1:54 PM
Subject: Re: [syndication] friendlier feeds


>
> Hmm, don't really even need to do that; using MultiViews, there are
> (at least) three resources available;
>
> http://www.example.com/feed <-- negotiated
> http://www.example.com/feed.rss <-- always the RSS
> http://www.example.com/feed.html <-- always the HTML
>
> Cheers,
>
>
> On Tue, Jan 01, 2002 at 01:49:00PM -0800, S. Mike Dierken wrote:
> > Accept is a great way to go.
> > When I wanted a non-html rendition to be used by a web browser, and
since it
> > is hard to control the Accept header from a web browser, I 'tunneled'
the
> > header in to a URL query term something like accept=text/xml (the 'do:'
> > means it is an emulation of HTTP concepts of the same name).
> >
> > There is some discussion here about this also:
> >  http://groups.yahoo.com/group/syndication/message/2633
> >
> > mike
> >
> > ----- Original Message -----
> > From: "Mark Nottingham" <mnot@mnot.net>
> > To: <syndication@yahoogroups.com>
> > Sent: Monday, December 31, 2001 12:25 PM
> > Subject: [syndication] friendlier feeds
> >
> >
> > >
> > > I've been playing with making feeds friendlier, using content
> > > negotiation.
> > >
> > > For example:
> > >   http://www.mnot.net/bookmarks/feed
> > >
> > > If you request it with a browser, you'll get an HTML page, telling
> > > you how to use RSS; if you request it with an aggregator, you'll get
> > > the RSS. This is because your browser sends an Accept: header
> > > containing 'text/html'.
> > >
> > > In Apache, I do this with this directive in access.conf:
> > >
> > >   <LocationMatch "/feed$">
> > >   Options +MultiViews
> > >   </LocationMatch>
> > >
> > > Then, whenever I publish a feed, I put in two files; feed.rss and
> > > feed.html.
> > >
> > > I haven't tested this with any aggregator except my own; if anyone
> > > has a problem, please tell me.
> > >
> > > The (slight) downside of this is that if your browser knows how to
> > > read RSS (or dispatch it to another program that does), it has to do
> > > one of two things;
> > >   a) know to put 'application/rss+xml' in the Accept header, with a
> > >      higher q value than 'text/html' (or just omit the Accept header)
> > >   b) pass the URI to the helper app, which makes the request
> > >      itself
> > >
> > > I think this is manageable.
> > >
> > > --
> > > Mark Nottingham
> > > http://www.mnot.net/
> > >
> > >
> > >
> > >
> > > Your use of Yahoo! Groups is subject to
http://docs.yahoo.com/info/terms/
> > >
> > >
> > >
> >
> >
> >
> > Your use of Yahoo! Groups is subject to
http://docs.yahoo.com/info/terms/
> >
> >
>
> --
> Mark Nottingham
> http://www.mnot.net/
>
>
>
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>
>