GPL'd code exists, but not GPL'd art [LWN.net]

GPL'd code exists, but not GPL'd art

Posted Mar 6, 2008 4:14 UTC (Thu) by emk (subscriber, #1128)
Parent article: Ryzom returns?

There are supposedly several GPL'd MMORPG engines out there. Here's a rather nice-looking
example based on Crystal Space:

http://www.planeshift.it/

Their art, though lovely, is effectively proprietary. So I think the real problem is getting
high-quality 3D models and graphics, not writing the code.

(Log in to post comments)

GPL'd code exists, but not GPL'd art

Posted Mar 6, 2008 9:43 UTC (Thu) by rwmj (subscriber, #5474) [Link]

GameBlender (a.k.a. Blender 2.0) is another. It makes it really easy to create interactive 3D games.

But you're right -- it's the artwork that matters, not the code. It's relatively easy to code up simple wireframe games if you have OpenGL, ODE and perhaps a game engine. Now try to get artists to contribute comprehensive artwork for free and under the GPL. This was my experience writing xracer.

Rich.

GPL'd code exists, but not GPL'd art

Posted Mar 6, 2008 10:16 UTC (Thu) by and (guest, #2883) [Link]

I'm pretty sceptic about planeshift. I have some (not much) sympathy about 
not releasing the artwork under a free license, but together with the fact 
that they require copyright assignments for code, it makes me think that 
they intend to take it proprietary once the quirks are fixed by the 
community.

GPL'd code exists, but not GPL'd art

Posted Mar 6, 2008 15:13 UTC (Thu) by tshow (subscriber, #6411) [Link]

Art is the key here. In any 3D game the art requirements are counted in man years, but
with MMORPGS it's much, much worse. There to put a half-decent world together the traditional
way, you need at least 10 man years of work from competent, dedicated artists. Beyond that,
you need continued artist input to keep the world going.

Speaking as a professional game developer, and one who has worked in (and run) startups,
it is very hard to find artists willing to contribute that much effort without a salary. Even
if you find one who is initially willing, the volume of produced results tend to be...
underwhelming. The art itself is usually high quality, but the rate of production starts low
and drops quickly.

Art isn't analogous to code here; while you can reuse artwork in other projects, there is
little value to be had in pulling it apart and tweaking the way there is with code. If you
know how to make a 3D model, there's not much that looking at someone else's 3D model will
teach you.

As a result, there isn't much of a "free art" movement the way there's a free software
movement.

Good art is a requirement for good games, with rare exceptions; you can occasionally get
away with procedurally generated artwork or simple art, but only rarely. To produce an Okami,
a Panzer Dragoon, a Katamari Damacy or a Rez, you need good artists dedicating a lot of effort
to the project.

Art goes beyond the basic visuals, as well. Artists are often responsible for a lot of
the level design and level metadata; if there are areas where a character can hide in shadows
or take cover, an artist has probably manually marked those. An artist may well have laid
down pathfinding information, collision data, material information (so you get the splash
sound instead of the crunch sound when someone steps in a puddle) and so forth. If a door
opens when your character approaches it, an artist probably laid down the trigger volume. The
artists are responsible for a sizable amount of the game experience.

If the free software community wants to be competitive in the game industry, the art
problem will have to be solved.

GPL'd code exists, but not GPL'd art

Posted Mar 6, 2008 20:53 UTC (Thu) by oak (guest, #2786) [Link]

> Art isn't analogous to code here; while you can reuse artwork in other 
projects, there is little value to be had in pulling it apart and tweaking 
the way there is with code.  If you know how to make a 3D model, there's 
not much that looking at someone else's 3D model will teach you.

Another reason for why it's not so easily reusable, is that the same 
graphics get old pretty soon. (how many games with penguins we have?)

New games need new graphics.

GPL'd code exists, but not GPL'd art

Posted Mar 13, 2008 19:04 UTC (Thu) by anton (guest, #25547) [Link]

I don't think that art is a problem in general, but it is a problem at the start.

Here's why I think that art is not a problem in general: For many (proprietary) games there exist free mods or levels, some with considerable artwork. As an example, take Grand Prix Legends (aka GPL:-), a proprietary racing car simulation from 1998, which was not even commercially very successful; the game itself came with 11 race tracks. For this game hundreds of tracks have been created, some looking much better than the original ones. There also exist lots of other add-ons for this game.

The license of such mods is usually not the GPL (if the authors have thought about licenses at all), but I think that the authors of the mods would release them under a free software license if that was the culture of the community (as it would be for a GPLed game). (There is also the issue of getting licenses for using the names and/or designs of real-world objects such as race tracks, but that's not a problem for all games).

I think the main problem for getting free art is that a free game has to become popular enough that there are a lot of players, some of which will start developing new art etc. in order to scratch an itch. For this to happen, the game has to be complete enough and interesting enough that it attracts an initial following, and for that it already needs some decent art. So the game has to be developed by a team containing programmers and artists with a common vision. And even then success is not guaranteed: lots of games try to become popular, and only some of them succeed.