« How to eat beef | Home | It’s too loud and I’m too old »
De-Babelizing Web 2.0 - a call for help RE: text markup systems
Posted by Mike Sugarbaker at 3:00 pm on 6/23/2005
(To my less techie readers, I’m sorry; this’ll just take a minute.)
So, Greasemonkey. You love it, I love it, it lets you hack just about danged everything. It lets people do lovely stuff like Instant Textile, which is a nice band-aid for when web sites ask you to submit text, but don’t do any nice auto-formatting for you, insisting on HTML markup instead. Sometimes I don’t wanna write HTML. Sometimes Textile is much nicer.
But! Textile is not the only attempt out there at being nicer than HTML. There’s BBCode, a curious tag set popular on forums like phpBB, which is basically HTML with square brackets instead of pointy ones. There are nearly as many kinds of Wiki markup as there are kinds of Wikis. And then there’s Markdown, which is Textile for people who hate some specific thing about Textile that Markdown does differently.
A Textile-to-HTML button on our textareas is lovely. Now, how about a universal damned translator? Something that figures out, from some contextual thing on the page, what kind of markup is required of us, and translates Textile (or anything else) to that? Yes. This is what we need IT CAN HAPPEN IN OUR LIFETIME.
Lingua Franca is a first (baby-)step. It converts a small subset of Textile markup to either HTML, BBCode, or MediaWiki markup. The translation is one-way, alas, and the version number here is 0.1, which means that most of it flat-out doesn’t work right. And the code’s ugly. And so is the UI. It’s doing everything right there in your browser’s little JavaScript-brain, and I am totally open to doing it on the server side of some server. In fact, I think that someone who provided a nice web service for translating back-and-forth between all these formats would earn glittering prizes… possibly on the Nobel level.
Anyway. It’s GPL. Hack it. Take it and run with it. Please. Or just leave a comment telling me about the brilliant stuff which I’ve just poorly duplicated. One way or another, let’s solve this damn problem.
June 24th, 2005 at 12:41 pm
ReStructured Text is a nice solution, and they already have good python code for parsing it into HTML and LaTeX and a well thought-out non-brain-dead structure.
June 24th, 2005 at 1:12 pm
ReStructured Text is nice, if a little under-specced for normal human use. I chose Textile because it has more traction, I’m familiar with it, Instant Textile was already there, and all sorts of other worse-is-better reasons.
If by “non-brain-dead structure” you mean that the Python code is already set up for easy, isolated coding of new parsing destinations, that’s great news. I kind of have my eye on XSLT, though.