Content Development - Need to think about framework

[Nye_Sigismund]

Moving on from last night, because I think that that's a good idea (and there's practicing what you preach, which is also a good idea) -

I've been thinking about how folks can organise open source content development. It's far trickier than code, because you're not dealing with tiny script files that are easy to read - you're dealing with art assets. So that was my question - how do you deal with that?

Cyan utilised a central hub Asset Manager. It's kind of sketchy what we know about it, but it involved the Max plugin frequently checking for updated assets on a central hub server. If we incorporate that into a sensible idea of how an age is made, at least graphically (props and textures made seperately from environments, then added in) we can begin to see how this sort of thing can be done.

When making an age, you have three principle groups of people:

1) Prop creators
2) Environment creators
3) Programmers

Development goes like this - everyone gets together to discuss a new project. It's decided on, and then what Cyan called Stage 1 happens - project outlining, concept, etc. Stage 2 is part of that outlining, with blocking out and general prelim work done. Then the project is reviewed (Stage 3), if passing moving on to actual development (Stage 4) and finally QA'd and reviewed again (Stage 5).

I think that if we take all of this, and adapt it in the following way, we'll see a pretty sweet development structure.

1) Open area for discussion of the project. Probably a forum topic. Concept art and the simple coordination goes here.
2) Some way for all the people who want to work on the project to communicate quickly. IRC and Skype are two examples.
3) Three seperate repositiories: Environments, Props, Code. Environments need to be build in a really co-ordinated fashion, although the page structure utilised by Cyan does make it pretty simple to hand out things for people to work on. Props having a seperate repo also makes sense because it means non-affiliated developers can clone that repo for their own use. Environments generally are not much help from project-to-project, except if you already have suitable textures done. Code having a seperate repo is pretty obvious.
4) An acknowledgement of "handover". Once the modellers are done and the age graphically completed, the programmers take over and complete the project, leaving the modellers to work on a new project.

Cyan termed this a "pipeline", and copying it does seem to me to be a really great way of getting stuff done.

I suppose that the elephant in the room is that there needs to be more than one developer who works on a project for any of this to matter, and I did get confronted with the argument that nobody really wants to work on common content projects these days a short while back.

So, two open questions:
1) Do you folks think that this rough idea, when fleshed out, could work? How could it be improved?
2) Is it feasible to expect more than one person to work on a project, given that this has been rare in the past?

EDIT: Because ages/levels are the main focus of content creation, I focussed on them here. But I think that looking into stuff like clothing and emotes is certainly important to. How can this be done?

EDITx2: Also, in before somebody links this on an IRC and then people start saying nasty things about me, which may or may not happen, but it is fun just to link this image:
http://3.bp.blogspot.com/_ZzXevucGsR4/S ... use_me.jpg

[Nye_Sigismund]

Having spent another day thinking about this - I'm really locked into the idea that this is actually a feasible way of doing things. I guess it's time to put my money where my mouth is and actually get this organised. No way of developing is ever perfect, and trying endlessly to polish the process just impedes that process.

Again, nothing locked in stone. But it's about time somebody actually decided to do this.

Okay, so, here's a content dev team outline, which I'm probably going to run with.

Code: Select all

 Team: "Open Source Creation and Restoration", or OSCAR for short.

 Mission statement: Develop new content for Myst Online: Uru Live Again and compatible fanshards.

 Who can join: Everyone. Even the people that may or may not hate my guts right now - not to be a whiner about this but it's worth putting here. No glass ceiling.

 The rules:
 Team direction is decided by consensus.
 Either leave politics at the door, or we'll throw you out of it.
 If you commit work, it's team-owned work.

If I do decide to work on this team group idea, I'll probably plonk down Llantern as a Stage 4 project. This is just so that I can get stuff onto the repos in question. Llantern's already got a JIRA and a Hg repo here, so that means I can get a big chunk of the organisation sorted quickly.

There's still the issue of if this is actually going to be of any interest to other people, of course. But if I do found a team which only ever has me on it, I won't stop. It'll just be a really small team.

Thoughts? I'll send PMs to the people I need to talk to over the next few days. Exams looming.

EDIT: After quite a while of searching, I came across an open source digital asset managment system that seems like it might be very handy to an open content team. http://www.razuna.org/ Free, and has version control, which is kinda important.