One way to solve it is to make people pass a quick programming test before they can buy a ticket

They were threatening to do that for last year's meeting, and then backed out on it. There was no test.

http://googlecode.blogspot.com/2011/11/google-io-2012-extend...

In the meantime, be sure to brush up on your coding skills. They’ll come in handy when the new application process opens in February. That’s all we can tell you for now, but we’d advise against making travel arrangements until then.

CAPTCHA: FizzBuzz edition.

this sure would make it a lot easier to get a ticket!

The current system is a lottery. If the page loads in the first couple of minutes your in, otherwise your not.

Well, there are a number of areas for potential improvement. I know that the absolute number of tickets sold isn't really going to change, and the price isn't something a startup could fix. I was more thinking about qualities like server load, customer fairness, etc. Google Play went down when the Nexus 4 launched, gave incorrect information to customers, spat out ridiculous errors, etc. And it seems there is no end to the complaints about the lack of notification or timing for very qualified and excited people to purchase tickets to I/O. A method of holding tickets until the purchaser demonstration of value, for example. Or something. I don't know.

Apple has the same problems.

* Increase ticket prices and then corporate people push out indie developers.

* Increase venue size and then each attendee gets less value out of it, and there's an upper limit to venues anyways.

* Add a lottery and now your most loyal customers may not be able to go.

It honestly seems like the best solution is to just throw open the floodgates at an announced time, and assume the people who really want a ticket will be the ones most committed to getting one.

Throwing open the floodgates at a random time might be an even better commitment filter.