[uf-discuss] Format-of-Formats?

[Phillip Pearson]

Re all this format-of-formats discussion, I should chime in with what 
we're doing (well, what we've *done*) in the Structured Blogging 
project, which is all about *publishing* microcontent...

--- PUBLISHING

Kimbro Staken designed an XML format called MCD, which is more or less a 
greatly simplified form of XSLT.  It's not a schema language, but rather 
a machine-readable set of instructions for editing and publishing types 
of microcontent.  The SB plugins (Wordpress and Movable Type) include 
libraries for interpreting the MCD files, generating editing interfaces 
and generating HTML and XML snippets to for publishing microcontent in 
web pages and XML feeds.

This is one piece of the puzzle.  Many may not like this approach as an 
MCD file defines only one editing interface and one way of displaying a 
piece of microcontent, when in reality there are many acceptable ways of 
editing and displaying it... but for the moment it's a very convenient 
way to bootstrap a publishing tool and make it trivial to support new 
formats as they arise.  Currently, if you develop a new microformat, the 
easiest way to allow your users to produce content marked up with it is 
to write an MCD file and send it to me to get it included in the 
Structured Blogging plugin distribution.  Of course there's no reason 
why you couldn't *also* make your own plugins that allow people to mark 
up their HTML in a more flexible way using your microformat, but 
building an MCD file helps keep everyone in the loop and bootstrap 
adoption of your format.

--- CONSUMPTION

It's important to mention (again) that MCD is all about *publishing* 
microcontent.  An MCD file doesn't tell you anything about what the 
fields mean or how you should consume the microcontent.  (Although if 
you also publish the XML that is used internally to connect the edit and 
display aspects, you can read this out of a feed or page and pull it 
back into an editing interface).

To consume the microcontent, you could extract the microformatted part 
of the HTML output if you have XMDP profiles for all the microformats 
used in the output (assuming the display part of the MCD file is marked 
up with microformats), or you could use the GRDDL hookup Danny Ayers 
worked out for us (it looks like the file got lost, but it should be at 
http://structuredblogging.org/subnode-to-rdf-interpreter.xsl).

Between MCD and GRDDL, RDF users at least have all they need to produce 
and consume microcontent.  I think we still have some work to do before 
this chain actually works properly, but it'll be there when someone 
wants it :-)

To do:

1. Ensure that SB plugins correctly add links to profile docs for 
microformats used into head/@profile attributes in output HTML.

2. Make sure each microformat has a profile document (XMDP or otherwise).

2. Ensure that all microformat profile docs include 
http://www.w3.org/2003/g/data-view in head/@profile and include a <link 
rel="transformation"/> element pointing to an XSL file that can turn the 
microformat into RDF.  See 
http://www.w3.org/2004/01/rdxh/spec#grddl-xhtml for details.

Danny Ayers has made an XSL file that turns the SB XML snippets straight 
into RDF, so that should work at the moment.  In future (after someone 
does items #2 and #3 above) it would be nice to replace this with a 
bunch of XSL files that turn the individual bits of microformatted 
content into RDF, in order to support microformatted content that's not 
published with the SB plugins.

Cheers,
Phil