There’s a discussion going on in John Blyberg’s blog about librarians as coders and how the right and proper direction for libraries is to have control over their catalogs so that they and their community can make improvements.
In it I hinted at some use of microformats as a way of improving catalog output, and the comment system there was dogged in rejecting my examples. So I’ll put it here.
The idea put briefly is that the RSS feeds and item displays in a library catalog or blog should include just enough semantic markup so that smart library clients – call them “patron applications” – can parse the data easily and make use of it. This might be some intermediate step toward the libraries themselves developing integrated services or it could be an end in itself to see what patrons come up with and run with it.
The specific suggestion is around the notion of microformats, very small bits of standards for XHTML markup that add some sematics to the markup while being still perfectly readable to anyone in an ordinary browser. And here is where I hope my code examples will work…
Rather than having a markup that looks like this:
<b>Title:</b> Anatomy of a Murder
that looks like this:
Title: Anatomy of a Murder
you have a bit of semantic goo that looks like this:
<b>Title:</b> <span class=”title”>Anatomy of a Murder</span>
which looks the same on screen, but which an intelligent feed reader or bibliography program or library bot or whatever you can imagine can read and repurpose to its heart’s content.
The only real tricky part is settling on the names of the tags, and agreeing that this kind of microformat isn’t going to have to be scrapped right away because of some grave problem, and being flexible enough with it at the start to do co-evolution of clients and servers. You would probably do well to start with Dublin Core for term names, and put together something incredibly small to start to prove that the whole exercise was worth the trouble.