Personally, I think it's better left to Mojo or someone else extremely knowledgable about jME best-practices to finish that tutorial as there are many ways to continue the tutorial, but I think one of the major benefits of the tutorial is learning the a good way of accomplishing the tasks in jME, not just one way that will work.
darkfrog
a good way of accomplishing the tasks in jME
No doubt. But on behalf of us who are still getting used to/learing jME documentation/tutorials are huge. The only complete set of documentation we have is Cep21's old guide, and a couple of old tutorials, that is still getting updated in the wiki. Contests may yield a great set of features but if we don't start documenting best practices we're falling into the stereotype that developers don't document anything. Plus we're hindering the noobs becoming pros. As more and more pro game houses consider jME having better documentation will speed its adoption.
I'm not going to keep whipping the documentation horse. It's topic has been thrown into the hat and even if a challenge doesn't come for it I'll still keep updating the wiki. Please just keep good comments in your source for the JavaDocs.
I do whole-heartedly agree with you. I only disagree that contests where this should be addressed. You raise very good points about documentation though and I think we'd all concur that it is something that needs to be handled. The only problem is figuring out how to handle it.
darkfrog
Fair enough.
The flag rush tutorial is the easiest jme specific thing to learn jme from but I don't think a contest would get those done (or maintain the aspects that make the flag rush tutorials so much better than the other available documentation, at least in my opinion). Maybe a contest to motivate mojo to finish it would be better…
Maybe instead of looking for contests that can improve the core jme stuff, maybe contests to produce tools that would be useful for developers. What are the areas that people see as the weakpoints of jME. I've always found the way models are handled somewhat of a pain, though generally it's the process of figuring out which models really can be loaded from what programs etc.
At this point I think a contest to determine the next contest might be the best thing to do…
Yes, I think so far we heard a lot of ideas about someone making something that we don't want to make ourselves. While I'd love to see most of those features too, is that a fun thing for a contest? Do we need several people to attempt to make a bone system? Damn, it's not even fun to judge 
I know several people have written
That's a good idea.
Perhaps we should decide on a category first and then decide specifics later?
What recommended categories do you all have? Lets not just throw out categories, feel free to post what category you would like to see a contest on next.
I've already tossed out AI, but I think that's been moderately shot down. 
darkfrog
I would like to see a contest to write a new JointController that can be understood fairly easily, flexible enough that when someone decides to write a model converter they don't have to write their own replacement class to animate their model (I assume this is a weights issue mostly but I don't know for sure) and finally, that someone could dynamically create a model from. I was looking at JointController and it looks like generally you would have to make a whole new controller to add one bone somewhere. It'd also be nice to have easy access to attach another model to a bone (part of the being able to dynamically create things here). From what I've seen short of doing a lot of work on it JointController is pretty much limitted to what is known at the creation of the model, you'd have to create a new joint controller to add anything to it, I think it should be almost as simple as adding a node to the scene graph…
Now, just in case no one knows, all my contest ideas will generally involve things I personally don't want to do or realize I'm extremely incompetant at the math involved (this applies to a LOT of 3D math, even if I can do the basic formula's easily, I never really understand how to apply them to get what I want hehe).
environmental effect? i'm in!! 8)
no monkeys for the developers! they already have one
(environmental challenge)++ from me.
And for an added bonus, the winner's doll will be exposed to a real world version of the environment they submitted before it ships ('cause sunburned monkies and soggy monkey are more fun…)
renanse said:
(environmental challenge)++ from me.
And for an added bonus, the winner's doll will be exposed to a real world version of the environment they submitted before it ships ('cause sunburned monkies and soggy monkey are more fun...) ;)
Just for that I'm going to push for lava effects. :P
We will have to decide what the minimum requirements for running the contest entries will have to be, graphics card wise. Since developing environmental effects just basically screams fragment shaders, one should think about if low quality fall-backs must be included or if we're going for the middle to high quality effects only. I'd say we try not to limit the participants and should refrain from forcing them on adding fallbacks, because hopefully the community could work on that on the winning entries, if we decide to add them to jME somehow (and somewhere).
If you make a really cool effect, that works on both older and newer cards, I think that deserves a little extra credit. But I don't think it has to be a requirment. Let's try and keep it open to as many different people and different ideas.
To me it seems like a lot of the work should be aimed at current high end/future medium range gamer systems… ideally things would be efficient (most things look better if they run really smooth anyway) but more realistic effects would probably be a better priority.
DanK, I would suggest that a key to good game development is being able to provide the best game experience with the least performance expectations. I think we can't really divorce one from the other even in a competition the same as we can't when we actually develop projects. The goal of these challenges is to provide features to jME and optimization is a key to being able to do so.
DanK said:
To me it seems like a lot of the work should be aimed at current high end/future medium range gamer systems... ideally things would be efficient (most things look better if they run really smooth anyway) but more realistic effects would probably be a better priority.
yes dude, just get me another graphics card and i'll totally agree with you XD
(environmental challenge)++
It would be nice if someone can code some terrain to start from. A terrainblock with some trees, rocks, (simple) water, … So it will be easier to concentrate on the environmental effect for the coder and the judge. (I hope you can understand my English )
How about we provide an ocean and island with some trees and it's up to you guys to come up with environmental effects for it? This provides for a lot of creativity as well as showing various results of the environmental effects…and lets us use water.