Paving the abstract road systems

Saturday 29th Oct 2011

So I read most of this, on the idea of replacing the ‘time’ tag with a ‘data’ one. So yeah, a bit of a rant, but it’s better than writing a tonne of tweets and being a bore on Twitter. So I’ll be a bore here.

I can understand why you’d want a nice generic tag to encode various numbers and whatever, but I don’t see the point in getting rid of a really useful and unambiguous tag like time. The arguments for replacing it seem to be about fairly specific use cases, dimensionless numbers, ranges of numbers, temperatures and so on. Fine, I can see that. It might be useful. What I object to is this sense of a ‘grand scheme’ creeping back in for HTML to become more like XML - a nice neat encoding of all the world’s data so that it can be easily indexed and rationalised by big systems. There’s much talk in that thread of microformats and machine indexing as if that was a major goal in making websites. I dunno, I figured it was about letting people read stuff.

What I liked about the development of HTML5 was the idea of ‘paving the cowpaths’ - looking at how people use HTML and either making that ‘official’ or adding stuff to make it easier. What we ended up with was a reasonably sensible set of tools that were nice and generic, salted with ones for a few specific use cases, like dates and times and form inputs or whatever. Yes, you’d be putting the tags in for a machine to understand them, but that machine would be a browser, something to present what you’d done - the need for marking up your content was obvious and understandable.

I liked the balance between neat and scruffy that HTML5 represented but this idea seems to be going too far back to the neat end, it smacks of XHTML2. You neats, you’ve got your XML, go and play with that, and let us be scruffy with our HTML.

Pave the damn cowpaths, stop planning highway systems in the sky.