Orbus Gameworks

OGDC Summary

May 14, 01:03 PM

I was in Seattle from Wednesday to Friday night attending the Online Game Development Conference. It was the first year that this conference was being put on, and holy smokes was I impressed at how well-attended and well-run it was!

First, I’ll mention the one thing I kept hearing over and over from other attendees: the food was fantastic. Much better than the usual crappy conference food.

But I don’t want the food to overshadow the excellent conference content. I didn’t get to attend too many talks, mostly because I was busy hanging around catching up with people I hadn’t seen in a while. But I’ll talk a little bit about the ones I did see:

Automating Game Balance. Christian Force and Nate Bogan of +7 Systems gave a talk on automated game balance for MMOs. I know the guys, since they’re Boston locals, but it pains me to say I felt their talk was maybe a little too over-general. I understand the need to not sell people things, and I never once mentioned the Orbus metrics product during my talk, but people need specifics. So I threw in a few things like, “for example, you could solve problem X like this,” basically citing the way we solve it at Orbus. Nobody minded, and people were happy to get some actual solutions laid out for them.

Software and the Concurrency Revolution. Herb Sutter is an old-skool C++ guru, currently with Microsoft. He gave a really fascinating technical discussion of concurrency, sort of a follow-up to his famous article “The Free Lunch Is Over.” Basically he was saying: when programming for concurrency, locks suck, but they’re all we have right now. But then he listed out a bunch of technologies and techniques that he thinks will reduce our dependency on locks greatly within the next four years. A little bit over my head, but he’s a dynamic speaker, so it worked for me.

Adventures in Middleware. Joe Ludwig, who’s the Director of Development on the MMO Pirates of the Burning Sea, gave a great talk on all the different kinds of middleware he’s evaluated, bought, and used over the last four years or so. Here are some tips he gave to developers who are evaluating middleware:

• Evaluate the company as well as the software. Will they be around in a year? Do they have sound engineering practices?
• Make sure to include artists and designers in your evaluation process. They might notice something they don’t like about the middleware that your programmers won’t.
•  Don’t be afraid to say no. Some people are tempted to say yes to a mediocre product because they’ve spent a whole month on evaluation and don’t want to throw that month away. But you’re not throwing the month away: you’re learning what you really need and don’t need. So, if you don’t love the product, say no!

Legal and Trade Ramifications of Selling Virtual Items to Players. Laaaaawwwww!!! The most interesting thing I got out of this presentation was that if your game world allows people to convert currency into and out of the game, you are sooner or later going to be considered a financial institution that needs federal regulation. Which means all sorts of overhead, largely due to problems with tracking money laundering and also insurance to avoid a run on your “bank.” Which means you’re probably doomed.

AutoAssault Postmortem. This was actually a really nice postmortem on what went right and wrong with AutoAssault. The developers did not bash their publisher. Their biggest takeaway was that they were given some rope and then hanged themselves: the publisher asked them to set milestone deadlines, so they did. And they technically made every milestone, but delivered sub-par product because they were too afraid to ask for an extension. Also, they said that one way to know your game is in trouble is if none of your employees play it. No matter what their excuse (“I’m too close to the game,” “it’s not my style”), if the game is fun, they’ll naturally be playing it. They’re actually enforcing a few hours of playing their new game each week. This also helps motivate bug fixes. If you tell a programmer “fix XYZ,” he’ll say, “Okay, later.” If the programmer gets frustrated by XYZ while playing, he’ll go in and fix it so he won’t have to deal with it next time he plays!

All in all, OGDC was an amazing conference, and I can’t wait to go back next year.

— DariusKazemi