Re: planetrdf.com - add Dublin Core feed?

OK, I'm getting there (with the attempt to XSLT-generate RSS1 from 
http://www.dublincore.org/).

Here's an XSLT (version tweak of Dan Connolly's one in 
http://www.w3.org/2000/08/w3c-synd), which turns a slightly amended 
version of the DCMI homepage into something close to the right form of 
RSS1. I'm cc:'ing www-archive@w3.org (a public mail archive) so the 
files can be pointed to later. This would all be easier if I actually 
knew XSLT :)

Attachments:

  dc.html (a tweaked version of the homepage)
  dc2rss1.xsl (altered version of DanC's XSLT)
  _out.rss (sample output, showing what's done and what's missing)

To make this work, the DCMI homepage needs some minor ammendments:

1. each news item needs a permalink pointer to its entry in the news 
archive (this is worthwhile anyway, so people can share links to DCMI 
news). In my sample, I've flagged this with a 'rel="details' attribute, 
to make it easy to extract the URI from that link. I added this to each 
item:

<div class="permalink">(<a rel="details" 
href="/news/2008/#dcmi-news-20080901-03">#permalink</a>)</div>

...but you could do anything really, so long as there is a predictable 
HTML markup containing the permalink URI that we need for the feed.


2. The date gets wrapped in <span class="date">2008-09-01</span> so we 
can extract it more easily.

There may be fashionable markup idioms for both of these (microformats 
etc. - eg. http://microformats.org/wiki/hatom), so perhaps we don't want 
to change anything yet. But then again these changes ought to be enough 
to autogenerate RSS1, which could save some time for you. If you care to 
make the edits in (1.) and (2.) I think we can have RSS1 generated 
fairly easily, although to put this into production would require having 
an XSLT engine in the pipeline somewhere. You could run this locally, or 
we could use a Web-based one such as that at W3C.

cheers,

Dan

--
http://danbri.org/