Stabs: Sure, having lots of people do the same things as you definitely increases the fun level, and sometimes you can only find a synergistic fun when you have a lot of people playing a game. However… I think that’s a lot more rare than you’d expect. I mean, imagine how much more popular EVE would be if they had taken the time to make the minute-to-minute details of mining and trading actually fun! (Yeah, I went there.)

Todd: that’s a good one for the list.

Darrin: I… I’m sorry to hear that :)

Fun is the main thing. Fancy graphics can catch your eye but to keep your attention it has to be fun.

Getting the game fun is always more important than a perfect infrastructure. You can’t make money if no one plays your game. But you can make *some* money if your server doesn’t support twice as many players, or if it has to be rebooted twice a day.

Publishers don’t care how cool your server tech is. Publishers give you money if they think the game will succeed. That is based on how fun the game is, and whether there is something they can see (or better yet, can play).

Content developers don’t care how cool your server tech is either. They care how easy it is to get content into the game at shippable quality. And there are a lot more content developers on a reasonable sized game team than anyone else.

The flip side is: build your server to be loosely coupled so you can replace parts that measurably don’t perform (there will be lots of “important” parts that just don’t show up on the critical path, and you probably don’t know the difference to begin with, so don’t optimize everything). Make your server scale-*able*, but not necessarily large-scale to begin with. Then, even if you don’ t have time to replace software, if you wind up hugely successful, you have the option of throwing hardware at the problem.

Have you ever thought about how the Portal team found the fun? Before or after the cool tech was finished?

BTW, Eric, I’m now working on a game that makes use of Smartfox. (I have just about eaten all the humble-pie in the fridge. :)

How about “version everything” as a corner not to cut? Then you can more easily fix your inevitable mistakes by bumping the version number. It’s better to have a few useless versions that never change scattered about than to get stuck with no way to fix or tweak some important element later on.

This is the core Catch-22. Only by having the experience of having done this many times can you know which things are critically important and which things are excessive. I completely agree with Eric’s statements here. In fact I’ll go even further, if in doubt assume you don’t need it. Don’t code any tech until the game forces you to make the new technology.

But having said all that, personally I just love coding tech more than designing games. ;) I’m not delusional enough to think that’s the path to success, but it sure is fun.

And Jason, my other advice would be don’t waste your time trying to make MMO middleware. The big players in the market like Hero and Big World are slowly failing after millions of dollars spent. Other companies are working their way towards bankruptcy trying to provide offerings for this market. The middleware that is surviving is the middleware that is more generic and supports a wide range of games. The MMO-specific middleware market is all dried up. Besides, the most successful middleware companies all started by making a game and then later realized others might be able to use their tech.

It’s the perfect essence of the two of you and a great way to re-imagine MMO combat. I like your niche so far. :)