This would be a good time to visit Montreal

I recently found out about Montreal Blitzweekend, a 48-hour programming/etc challenge on November 24 and 25, 2007. macournoyer from Standoutjobs has an interesting post urging you not to participate.

I personally love this sort of thing. As I said about last November's NaNoWriMo: Impossible deadlines and absolutely no quality standards? Hello, count me in! (Incidentally, I'm doing NaNoWriMo again this year too.)

Which all reminds me of the RapidPrototypingContests (RPCs) that we used to do at NITI. These were challenges where we'd break into teams to see who could build the best prototype of the given requirements in a single weekend. Mostly I was the one who wrote the rules, and I mostly ended up judging the results. So perhaps the best thing about Blitzweekend compared to an RPC is that I can participate and I won't have to hold myself back when I do :)

A note on naming: Blitzweekend vs. RPC

There are two points in the name "Rapid Prototyping Contest" that do not come out in "Blitz Weekend," giving us a hint that it's targeted a bit differently. The points are:

1. Prototyping. "Blitz" implies producing something useful in a short time. Prototyping implies *trying out* ideas in a short time. Trust me, you will not produce a valuable product in 48 hours. That's by definition: if it can be done in 48 hours, it can be done by your competitors in 48 hours, so it's not valuable. At best, you might produce something that *looks* to other people like it might be a valuable product. But you, the developer, have to see it for what it is. Most people are very bad at letting go and admitting they're working on a prototype. And almost everyone is bad at throwing away their prototype and starting over from scratch to build the real thing.
2. Contest. I have no idea of whether BlitzWeekend will be competitive - ie. with rules and judges - or not. I suspect not. In fact, the most fun RPCs we had were also not very competitive, and it saves time on judging, so that works for me.

And yet... How to win at Blitzweekend ...and yet there's always a winner, whether it's official or not. The winning team is the one who makes the obviously best product. Naturally I intend to be on the winning team. macournoyer gives three good tips on how to be successful in the Blitzweekend and in general. Here's my tip: whatever it is you're doing, do it along with awesome people who have done it before. It so happens that my fellow alumnits have done it before - four times! In RPCs, we competed to make working VoIP client-server applications in 48 hours (before SIP was popular); desktop search engines in 24 hours (before there was Beagle); ProjectDeathRay (an integration-testing framework for operating systems) in about a week; and some amazing new web-based configuration interfaces in about 36 hours. Who can imagine what'll be next?