GDC 2009: Making LOVE In Your Bedroom
"This is the Game Developer's Conference - this is not E3 - so my plan is to not actually talk too much about my game. I'm actually going to talk about the process of making a game," but Eskil Steenberg still pulled a few trees out of the ground, and ran around a bit shooting some energy bullets in his procedurally generated, user modified world, LOVE, the MMO he is single-handedly crafting.
It's true, though, that the majority of his presentation was focused on tools - tools that he created himself. Moore's Law is pushing processing speed (and hence graphics) so fast that Eskil has proposed his own law of game development, "Unless you can produce twice the amount of content per man hour every 18 months, you're screwed." You can't just hire people, so you build tools. Everyone focuses on solving the hard problems, but actually the vast majority of your time is spent on pretty mundane things, so if you can get some really good flexible tools going for your everyday activities, it can make your game developing life a lot easier.
Eskil's slim team philosophy extends to probably firing your designer. "You have 199 people trying to get stuff done and one sucker trying to get more stuff into the pipeline. He's trying to make that to-do list longer," to which he adds, "To me, hiring a designer who can't program or do art is a little bit like hiring a personal shopper with a great sense of fashion but doesn't have any value of money. So you go and say, 'Here, go buy an outfit,' and they come back with a pair of socks like, 'Oops this is all I got. Sorry, ran out of money,' and they might be really nice socks, but you know, it's not an outfit." Basically, make all your hires multi-purpose. Don't hire a single uncreative person.
He also thinks you can do away with concept art. Why not make your tools such that by the time you have conceptualized your object, it's also modeled? That's precisely how he works, and his tools are all hooked up to a Verse server (as is LOVE) so that when he putzes with an object via his tool, it is affected immediately in-game. Can you imagine working on a network like that where there is no loading the latest build?
Things got pretty technical, but if you're so inclined, it's all open source. Eskil is the kind of friendly guy who encourages you to take his tools (one-handed -- so that, like Pong, you can operate them properly with a girl or a beer in the other) and be productive, tweak them to your needs, and make cool games. He doesn't have one for texturing yet because, as for that problem, "In my game, I was smart enough not to have texturing, so...solved!"
Eskil? Sounds swedish. :)
This is a pretty cool evolution of the development process. And I like the painted feel of the rendering engine; even primitive models look artistic and interesting. Could be a big wake up call for major developers to try something new, if this pans out.
Sounds like an idiot to me honestly.
If you throw the designers and the concept artists out then you are going to have one bland, uncreative and boring product.
Just look at Grim Fandango or Parappa the Rapper, the art style is phenomenal and it doesn't inhibit the graphics in the slightest. If they had thrown away the designers than the quality of the games would have suffered a lot.
I say: Hire more people designing and drawing ideas and get rid of part of the team that revolves around making the graphic engine. Who the fuck cares about graphics?
Design is not optional, unnecessary, or disposable. It is integral-- everything else in the game revolves around the design! Programmers, artists... whatever content or technology is used is developed in accord to the project's design goals. Sure, designers don't make it easy, but that's their fucking job. They have to make sure things are cohesive, continuous, and complete. That is their skill.
Don't like it? Get over it. The designer is there for a reason, and the fact that you answer to him is the essence of it. The designer knows what he's doing, what's going on, and what should go in or be cut out. If he doesn't, you should get a better one, but NEVER consider getting rid of designers altogether!
I mean, anyone even considering an MMO without heavy design work is an idiot. MMOs require TONS of design, and even with 3rd party software, outsourcing, and tools, it is certainly more than reasonable for their to be someone, or even several people, who are just working on the design and making sure everything ticks. Consider all the changes to gameplay made in patches to MMOs-- that's designers hard at work. Think WoW would have gotten this far without its expert designers working the polish?
I would actually go so far to say that games need better design. Does that mean more design or designers? Not necessarily, but possibly. People bitch about games being juvenile or cheap. All that has to do with the quality of the design.
And, Steenberg fails for saying designers are uncreative.
That guy is underrating game design. It's very important to be able to estimate the impact of any design choice and know a bad idea before it gets implemented. Having ideas sounds like something anybody can do but you want GOOD ideas and those require experience and skill to filter for (as many, MANY failed attempts at innovation demonstrate every year and a bad design is almost impossible to salvage). Of course you can develop those on an existing team member (may be expensive though, experience takes some tries to build up and those tries are going to cost a ton of money) but don't think that a designer is just a monkey in a suit who tells you the obvious.
Concepts are done in an easier to change medium so you can alter them again before implementing stuff.
Of course it's possible to "wing" it, that's what agile development is designed around but as easy as that sounds agile development requires STRICT discipline, if you mess up any part of the rules you end up with an unworkable trainwreck. If you can keep that discipline up (which, among other factors, includes NO OVERTIME), good but if you can't it's game over for you.
Of course it's much easier to do agile development when you're working alone, all software design principles are geared around coordinating a team to get the job done, a lone person doesn't need coordinating.
What you DON'T want is a team where every team member gets to make design choices, sounds great on paper but in practice you'll end up with something between a feature katamari (picks up any ideas that are smaller than itself) and a straight clone of whatever game the team loved the most. That's why opensource game development tends to fall flat, the team will either want to clone something or throw in every random feature they can think of (and often both at the same time!) with no single person who gets to decide what goes in and what not. I've seen that in action, it's not pretty.
Yeah, skipping texturing is a massive time and motivation saver, the alternative is to think of a way to make your textures require minimal effort (e.g. stylization). The problem is that textures are a computationally cheap way to add detail to a surface so skipping them can either impact your performance or just make your game look dull.
I tend to agree with the man in some respects, provided you take it less of a "Throw the designer in the skip and set fire to it" and more of a think less about hiring a designer and a programmer and more of a programmer/designer. "Duh" you might say, but to me it seems more like a slight at programmers as much as it is at Designers.
Too many programmers just take design in one end and crap out the resultant code. As good as that code may be, a smart programmer who pipes up that perhaps doing this or that differently would be a far better employee. Not in terms of the complete design proces, but you just have to look at games like WoW where the programmers clearly just do as instructed, because there's parts of that game which are easily improved but are left alone because clearly their "design" team isn't too bothered by little niggles (Armor dyes etc.) which could be fixed by the programmer if he was allowed to think.
But of course, I don't have any experience as part of a functioning games development team, so I could be wrong as hell, but provided the programmers are kept on a leash, what's wrong with a bit of design/program hybrids?