POWDER: Not So Quiet

Since it completed the Recommendation Track process last year, little has been said or written about POWDER. However, there have been a number of unrelated events recently that I take as evidence of a long term future. As chair of the WG that created it (and an editor of all but one of the documents and general front-person for the whole thing), this makes me happy!

One of my private measures of success for it has always been that one day, someone I don't know and who doesn't know me stands up at a big conference and says "you know this POWDER thing is really cool." That happened at SemTech last month when Matt Fisher presented it in a session called RDF Friday Part 3: Practical RDF - POWDER & Object Design Patterns. The full version of what he was saying is available in an article on his company website Putting POWDER to Work. Matt and I have exchanged e-mails since then but we hadn't had any contact before.

Secondly my friend and WG member Andrea Perego has been cooking up some code that uses POWDER to generate RDFa in a way that could make it really easy to add all those <link /> elements in documents on the fly under the control of a single, central POWDER file.

Suppose you want to add RDFa to all the pages on your Web site (not a bad thing to do!). One can imagine doing this for Creative Commons licences, DC metadata etc. Something like

<link rel ="cc:license" href="http://creativecommons.org/licenses/by-nc-nd/3.0/" />
<link rel="dcterms:creator" href="http://philarcher.org/foaf.rdf#me" />

These link elements should probably be included on every page of your site. Sounds like a job for POWDER. Andrea's PHP POWDER Processor (3P) can take a POWDER file and URI as inputs and return those RDFa link elements via a RESTful API - one that could easily be called from within an authoring tool. Full documentation, including an example using the Open Graph Protocol, is available.

Another development is still under wraps at the moment but the signs are very positive that a combination of marketing expertise, industry contacts, business dynamism and, not unimportantly, venture capital is coming together in a POWDER-fuelled start-up.

A quick reminder of the key features of POWDER:

• it allows you to associate a bunch of predicates and objects with any number of subjects (as Dan Brickley puts it: it solves the aboutEachPrefix issue);
• it's primary format is XML, a small amount of which can describe a large amount of content;
• it has an associated GRDDL transform that renders the data as semantically-equivalent OWL;
• a POWDER processor returns RDF triples;
• the provenance of the data is always declared.

If you haven't looked at POWDER before, maybe now's a good time to do so.