28 October 2010

The Limits of Openness?

I've been a long-time fan of the 3D modelling program Blender. No surprise, then, that I've also been delighted to see the Blender Foundation moving into content production to show what the software can do.

Specifically, it has produced a game (Yo! Frankie) and three animated films: Elephants Dream; Big Buck Bunny; and most recently, Sintel. Aside from their aesthetic value, what's interesting about these films is that the content is released under a cc licence.

Here's a fascinating interview with Ton Roosendaal, head of the Blender Institute, leader of Blender development, and producer of Sintel. It's well-worth reading, but there was one section that really caught my eye:

we keep most of our content closed until release. I’m a firm believer in establishing protective creative processes. In contrast to developers — who can function well individually online — an artist really needs daily and in-person feedback and stimulation.

We’ve done this now four times (three films and one game) and it’s amazing how teams grow in due time. But during this process they’re very vulnerable too. If you followed the blog you may have seen that we had quite harsh criticism on posting our progress work. If you’re in the middle of a process, you see the improvements. Online you only see the failures.

The cool thing is that a lot of tests and progress can be followed now perfectly and it suddenly makes more sense I think. Another complex factor for opening up a creative process is that people are also quite inexperienced when they join a project. You want to give them a learning curve and not hear all the time from our audience that it sucks. Not that it was that bad! But one bad criticism can ruin a day.

Those are reasonable, if not killer, arguments. But his last point is pretty inarguable:

One last thing on the “open svn” point: in theory it could work, if we would open up everything 100% from scratch. That then will give an audience a better picture of progress and growth. We did that for our game project and it was suited quite well for it. For film… most of our audience wants to get surprised more, not know the script, the dialogs, the twists. Film is more ‘art’ than games, in that respect.

That's fair: there's no real element of suspense for code, or even games, as he points out. So this suggest for certain projects like these free content films, openness may be something that needs limiting in this way, purely for the end-users' benefit.

Follow me @glynmoody on Twitter or identi.ca.


Nick Holden said...

Just a quick observation, but surely if it is "for the end users' benefit" then the decision can be left to the end users as well?

I think it is likely that even if Sintel had been available through some kind of 'nightly build' svn, I would have chosen to wait until the creative people putting it together were happy to call it 'released' before watching it, just to get the full experience on first watching.

But that doesn't necessarily mean that I should be prevented from viewing a development version of a creative or artistic product if I choose to do so.

For most people, the 'buzz' of being able to do so would probably wear off soon after such an open source art model became commonplace, but for the few who would take the trouble to download repeated versions of the same artistic product I can't find it very likely that their critiques would be harsh. If you're going to go to the bother of watching ten versions of the same film, you probably care enough to give constructive feedback.

Glyn Moody said...

@Nick: I think the problem here is hearing/seeing spoilers inadvertently. People are likely to mention this on Twitter/Facebook etc., or maybe in the office: if the plot is available, it will surely come to the ears of those who don't want to hear it...

Maybe the solution is to make staged versions available *after* release so that people can see it evolve...