Microformats from a programmer 12 Sep 2005

If you’re interested in structured data or writing parsers for commonly occuring publishments on the web like blog posts, calendars, contact details and so on, then you should check out microformats. Specifically, download the MP3 of the microformats discussion at BarCamp. It serves as a great introduction to mircroformats from two people that really know what they’re talking about, Ryan King and Kevin Marks.

Even though I’d heard of microformats before, I still think I learned a lot from listening through the 70 minutes of audio. They talked a lot about the principles for microformat design, which make a lot of sense when explained well. Microformats are a good idea from a producer/publishers point of view because they’re designed to be easy to read and write, and to fit into the already existing software.

However, I’m not really a producer or publisher; I definately fit into the programmer category. What makes microformats really cool for programmers, though, is the fact that they’re standardising problems that every blogger solves every day in his or her own way. They’re saying (for example), ‘If you’re marking up a series of events, do it this way’. This convergence of implicit structure built on convention to explicit structure built on a formal specification is really exciting. If support for microformats really gets rolling, that means that everyone is marking up their events in the same way. As microformats are also designed to be machine-parseable, that means that everyone’s events are suddenly machine readable and, more excitingly, aggregatable. Just think through the cool things that could be built on a tool that crawls web pages for hCalendar snippets and aggregates them into a centralised list. And that’s just a single microformat. There are so many more. Score one for the openness of the web as a generic publishing platform.

You probably don’t know, but I’m working on a generic compound microformat parser, Comfort. The current release does fairly well at parsing the basic hCalendar spat out by Comfort’s parent project, bCal, however I’m working on it to be a more useful parser. I’ll be throwing a lot of data at it because I want this to grow into a really powerful parser. Microformats aren’t exactly easy to parse (as they’re designed for humans first, all the trade-offs and comprimises that are a natural part of specification design make microformats that bit more difficult to parse), but I think there’s a real need for a generalised compound microformat parser. It would have to be easy to extend, as the microformat project is and hopefully will continue to grow as unbelievable rates.

If you’ve never seen microformats before, do check out microformats.org and specifically read the introduction on their wiki. Whether you’re just Average Joe blogger or are interested in building tools that can pull from the biggest database of all, microformats should be the way of the future. Lets just hope they aren’t crushed by the corporate weight of the W3C and its Semantic Web.