Contest Entry

Well, we got one entry (ehubbard), which isn't terrible considering it's the first. But, with one entry there's not a whole lot of competition for him. So, I say we evaluate his entry and determine if it met the requirements of the contest and is worthy of a couple prizes.

If we get back with an answer before the end of the week, that'd be good.

one entry :-o - that's quite poor :expressionless:

mojomonk said:

Thanks to all the entries


have not looked at the code yet(im off to a wedding for my best friend), but i really think we should give credit for daring to enter and send in your code, even if it's pretty shitty…at least i can tell from his posts that he has put some time in it…

i'll do some benchmarking with it and compare to naivevegetation, hashmapveg. and quadtreeveg. just for fun…

darkfrog, where's your entry? didnt you want to win that t-shirt with "i crushed MrCoder"?

even though we didnt get many submissions i still think it has gone pretty well…got alot of really interesting discussions in the thread, and close to 200 replies :slight_smile:

hope this tradition expands in the next challenge…i for one learned alot on how challenges should be made and how hard it is to find an appropriate challenge topic + rules…

I think more people wanted to participate (I was even working on something myself), but were put off by the deadline. It's not that it's impossible to make an entry in the time that was given, but… it does leave little room for delays, extra discussion, etc. If you judge it by the "buzz" it generated it was reasonably succesful, but by te entries, like irrisor puts it, quite poor. I think, establishing a deadline while more or less the rules were still being set didn't help much either.

I assume in the future we'd rather have a few contest with more participants, than many contests with a few participants… and I think this contest gave us some clues on how we could achieve that. Maybe for the next contest we can open a discussion first, on what the subject of the next contest should be, to see what the areas possible participants are intrested in.

I'll also look at the one entry as soon as I can get my workspace fixed again (the joy of upgrading Eclipse…). If it displays the vegetation, I guess it deserves a price in this case :slight_smile:

@MrCoder, I figured out it was easier to print my own shirt that says it. :wink:

Actually, I did want to contribute to the challenge but the timing was poor for me as I'm in the middle of so many other things I just couldn't find the time to work on it.

To that end, I would suggest considering extending the deadline by a month or at least asking if other people were wanting to participate but the timeline was too much?  I know it might give me an opportunity to contribute something and hopefully all the other people that were interested might be able to get the time to put something together as well…

Just a thought. :wink:


this was a trial/test challenge after all…but i really agree that it would be great to have some discussion around the challenge topic before kicking off the challenge+deadline…lots of things i would have wanted to change in the challenge after seeing everyones posts etc.

bedtime now(just got back from the wedding party) phuuu…

I'd say let's close this one, determine if ehubbard did enough to "win" and move on to the next one.  We'll look more official that way as opposed to reopening the contest after saying it was over.

a good point renanse.

What do you guys think about coming up with a more directed challenge for the next one?  Like a contest to make the best tutorial game (sort of like on the wiki for FlagRush) or maybe a clone of a retro game?  This way we can get some extremely beneficial promotion of writing actual games.  I think it would be awesome to have a 3D pacman, or Asteroids, or something old-school that could be re-done a thousand different ways and judge for creativity, enjoyment, features, etc and then we can make webstart links available for all of the contestants and start a poll of the people in the forum to determine the winner?

Just a thought - I've always got them, whether they're good or bad is up to others to decide. :wink:


hmm, I expected something like this:

sfera said:

how many submissions did you get mojo?

Will you answer? I'm afraid the good 'official' look is gone then

a clone of a simple gameplay retro game would be cool

it would be cool!

But would you expect better participation? I doubt it because it requires much time to make a full game - even if it's really a small game.

true…that's why i made a challenge-framework that takes care of the boring stuff(though it wasnt all great)…i like when things are made like all bot-games out there…fill in your ai in this class using these methods and then let your bots fight for survival etc :slight_smile:

renanse said:

It might not run the same each time if we had traps and random arena events the AI had to deal with.  :)

Bwahahaha! :o

We could push at them for the example stuff the "easy levels" which would give them relatively basic opponents that don't do a whole lot more than the basics, and the actual testing would consist of various degrees of increasing complication being imposed as they combat to be able to determine how "smart" the AI is and how able to adapt to surroundings.  I like it.

Any ideas on what type of game?  A robot style arena fighting game would be my first thought (sort of like, but probably quite a bit more simplistic.  We can try to talk Mr. Coder into doing some really cool death explosion effects for us as well. ;)


hehe that would be fun s start game->die->end, start game->die->end… :wink:

lets instead write the pacman gameplay and let people try to do the best looking game :slight_smile:

anyone tried ehubbards code? on my comp i get marginally better fps than the naive implementation(much faster startuptime though)…

I haven't yet, but will try to do so tonight if I think about it.  I'm chasing multithreading issues in JGN 2 at the moment. :-p

I would think "best looking" would be less helpful to the jME community though.  Something we're definitely lacking around here is AI support in my opinion and the more we can get people to be thinking about AI the better it will be for the community to get some developments there.

What about a pacman implementation that has basic ghost AI and they have to implement the pacman AI that will beat the level? :slight_smile:


I also had an AI based idea… We could provide a battlebots type framework for the next contest and people could program in some AI to see who could build the smartest, most lethal bot.  Mark pointed out to me this morning that IBM has done this already though, and plus I'm not sure if we'd end up with much that could be reused by jME… still, I think we'd get a lot of participation if the API was pretty simple to use.

Actually, that's a really cool idea, and when we join them all in the arena we could record a video of it and release them on the web site. :slight_smile:

Each bot gets three rounds maybe?  Although, most AIs would probably end up with the exact same results each time.  We could have a one-on-one tournament scenario and each bot ends up climbing the ladder.  This way there would be a winner, but it would also show how well each bot stacked up.

We could collaborate to write the base arena and bot interface that would receive events and have a specific API that they could work with to find out information about their surroundings and execute certain actions.  We simply release an API that gives them access and perhaps a few demo implementations they could face up against for testing.


It might not run the same each time if we had traps and random arena events the AI had to deal with.  :slight_smile:

Okay, I've got it then…we go ahead and write "retro pacman" and leave out the AI implementation of the ghosts. :)  The contest is to make the AI extremely competent so pacman can't win. :slight_smile:


I purchased the t-shirt today. Might as well announce that ehubbard won. Since the rules stated that it had to be better than NaiveVegetation, not Mr. Coder's Quad or HashMaps, he meets that standard.