LWJGL16k

[swpalmer]

I'm very annoyed that 1.5 isn't on Mac yet, as I don't have a Java high resolution timer and no LWJGL either .

$500 gets you Tiger (OS X 10.4) Betas + final version, ADC membership for 1 year, developer tutorial DVDs, and Java 5 betas for Mac  -- right now.

While I admit I haven't given them my $500, it does seem like a reasonable offer, considering the OS upgrade alone will be about $130.

LWJGL is working fine on the Mac today with Java 1.4.   I've played the latest from Puppy Games and more and there are no major issues with them.

As for the LWJGL competition,  I think it is a great idea... but what I would like to see more of is Java entries in other game programming contests that are NOT directed a Java.

The Mac community has an annual contest for indie gamers and I'm sure that Puppy Games and Odd Labs could score big there.

The idea is to show the larger game programming community that Java really is ready for mainstream games.  We aren't about cute little applets anymore.  Once we show off a few star titles in places where they had to compete against the traditional gaming languages the Java gaming position will be that much stronger.

[William Denniss]

ah yes, but my users won't be shelling out $500 bucks just for Java alphas to play my game

When Java 1.5 is released for Mac, will it be available on 10.3 as well?  I hope so.

LWJGL works great on Mac, I am using it as my renderer of choice now for Xith3D.  What I meant is that one can either use Java 1.5 or LWJGL to get a standard high-res timer.  Alas, for the 4k compo I can't use either.  In the end it actually didn't matter, since I am using asynchronous draw updates anyway (which is cheaper bytewise).

Will.

[swpalmer]

When Java 1.5 is released for Mac, will it be available on 10.3 as well?  I hope so.

Apple hasn't said.  It may happen, but I wouldn't bet on it.  My vote is that OS X 10.4 is required.  Apple makes Java an integrated component of the OS.  Sometimes Java problems are solved with OS updates.  E.g. I wouldn't be surprised if they added hooks to the window manager to handle openGL acceleration in AWT better or something like that.

LWJGL works great on Mac, I am using it as my renderer of choice now for Xith3D.  What I meant is that one can either use Java 1.5 or LWJGL to get a standard high-res timer.  Alas, for the 4k compo I can't use either.

Ah, I see.

Games published by our own members! Go get 'em!

[tom]

The 4k compo is over and it was a great success. Is it time to start a 16k compo?

What will be the rules/deadline?

It's important that to get as meny entries as possible. How about allowing any major java game lib to be used?(Jme, xith, jogl, jaol, lwjgl, java2d, javax.vecmath, ode, etc)

[woogley]

The 4k compo is over and it was a great success. Is it time to start a 16k compo?

What will be the rules/deadline?

It's important that to get as meny entries as possible. How about allowing any major java game lib to be used?(Jme, xith, jogl, jaol, lwjgl, java2d, javax.vecmath, ode, etc)

First off.. the 4K compo is still being judged, so hold your horses

I'm planning on hosting the 16k much like I did for the 4k.. only not nearly as "last minute". I'm working up a site right now as we speak, I think I can have it up by the end of March or even earlier than that. I say go ahead and start coding.. and the contest should begin in April.. and give the coders 3months to do their stuff  (which means it would end at the end of June). Any objections to this? How about some rules? I say the 16K should be freestyle like the 4K.. so lets not limit it to LWJGL! Someone correct me if I'm wrong.. but if the end-user has LWJGL installed on their machine, the game does not need to package its own LWJGL, correct? This can mean a full game in 16K without having to worry about the bytes taken by packaging LWJGL with the game. We need to settle on the target versions of the common libraries (LWJGL, JOGL, GAGETimer, etc.). I think that should be the first order of business.

[blahblahblahh]

The 4k compo is over and it was a great success. Is it time to start a 16k compo?

What will be the rules/deadline?

We had a poll that went fishing for the various sensible rules, and I started work on a 64k + libs comp with a 2-week time-limit. Although, given the success of J4k this year, perhaps a 2-month time-limit would be better?

I've secured some commercial assets for use in the comp (sound effects etc), but didn't want to start anything until the dust from j4k had settled. The main thing is to be careful we don't have lots on the go at once, because that's likely to spread the attention too thinly .

Given that GDC starts next week, I'd rather wait till it's finished (and see whether GTG announces a new comp) before starting anything else.

[dsellars]

The 4k comp this year has been a great competition.  It even got me to post an entry  

Looking at recent posts it seems that a number of peope are wanting to to go a head with a larger scale competition, which is great, I might even take part again.  

I'm just a bit concerned that in a month or two we'll have a number of different competitions that are being run by different people, with slightly differnt constraints at the same time.

Are you guy's talking to each other about whats to do next?  It's just that otherwise we'll end up spreading entries between the comps too thinly.  

The good thing about he 4k one was that most people in this community gathered round and focused on it (some more than others )  If we have a number of differnt competitions then it will be to the detriment of all of them.

I hope this makes sense and that I havn't read the situataion wrong.

Regards,
Dan.
       

[princec]

If we could have the LWJGL, DevIL, JOGL, JOAL, JInput, and SDL libs specified as extensions and hosted, then we could mandate that the entry is launched via webstart and references the specified extensions. The size restriction is on any non-library jar files, excluding the JNLP file. Sound ok?

Cas

[blahblahblahh]

but didn't want to start anything until the dust from j4k had settled. The main thing is to be careful we don't have lots on the go at once, because that's likely to spread the attention too thinly .

I'm just a bit concerned that in a month or two we'll have a number of different competitions that are being run by different people, with slightly differnt constraints at the same time.

[oNyx]

>Sound ok?

Ye.

P200/gzip allowed? Because that would make a *big* difference especially if you have a buch of classes. Effectively 25-35% more room.

Games published by our own members! Go get 'em!

[blahblahblahh]

If we could have the LWJGL, DevIL, JOGL, JOAL, JInput, and SDL libs specified as extensions and hosted, then we could mandate that the entry is launched via webstart and references the specified extensions. The size restriction is on any non-library jar files, excluding the JNLP file. Sound ok?

Yeah, that's pretty much exactly what I'd been tinkering with - the "upload new version" system on JGF is pluggable with different handlers for different "releases", so ... I've (locally) got a "64kcompetition" release, which looks through the uploaded game, checks total filesizes etc, and generally verifies it's a "valid" competition entry.

Then, because it's now tagged, that particular release of that game then shows up on the competition's page as one of the releases for that competition.

I'm currently working on the "automatically assign an entry to a particular "round" of the competition based upon submission timestamp", but not got anything working there yet.

...although first I want to get the "hosted JNLP extensions" running. Worst-case scenario, if it's taking a long time to get ready: in one afternoon I can manually put them on JGF, customize the HTML template that builds the submission form to have "use LWJGL?" etc checkboxes, and the JNLP servlet will pick that up and use my (hard-coded) local URL's to the extensions.

But, assuming no comp is going to start in the next 5 days, I'm currently taking the time to make a general-purpose "extensions / libraries upload system" so all the other games on JGF can share refs...

[princec]

I'm afraid we shouldn't allow pack200 as it excludes the Mac which is crucial.

Cas

[dsellars]

on Today at 9:12am, blahblahblahh wrote:
but didn't want to start anything until the dust from j4k had settled. The main thing is to be careful we don't have lots on the go at once, because that's likely to spread the attention too thinly .


on Today at 9:48am, dsellars wrote:
I'm just a bit concerned that in a month or two we'll have a number of different competitions that are being run by different people, with slightly differnt constraints at the same time.  

That'll teach me to skim read before posting  

[oNyx]

> I'm afraid we shouldn't allow pack200 as it excludes the Mac which is crucial.

Well, that's alright methinks... I mean... 16k that's a cubic s***load of space

I just wanted to have that issue clearly defined for avoiding confusion. Since I wasn't able to finish my second 4k game that issue was avoided for j4k. (Maybe next year's j4k should be clear in that regard, too)

And as suggested before some kind of readme might be nice, which would also clear the redistribution rights (and who made it, the website, what the game is about etc).

Here is a draft:
http://kaioa.com/k/4kreadmedraft.txt

[William Denniss]

I'm afraid we shouldn't allow pack200 as it excludes the Mac which is crucial.

Cas

Agreed, at least until OS X 10.4 comes out.

I've found that even just using a good zip compressor saves 10% or more space!  The default ones are pretty crap.  7zip has a very good standard-zip compressor if you tweak it with the right command line options.

Will.

[Orangy Tang]

We had a poll that went fishing for the various sensible rules, and I started work on a 64k + libs comp with a 2-week time-limit. Although, given the success of J4k this year, perhaps a 2-month time-limit would be better?

Hmm, 2 months is a bit too long I think. Anywhere between 2 weeks to a month is reasonable. Anything over a month and you've got to start actually thinking about stuff rather than just hacking it as you go.

Given how much people get done in the LD48h compos, 2 weeks should be plenty. So a month gives a chance for multiple entries.

[oNyx]

Well, this year's j4k had also a pretty long run. I think that was ok, because it allowed more people to participate. Most people just can't afford to spend 2 weeks straight on a (just for fun) game, because they lack spare time in such a big chunk (which incidentally comes up at the compo time).

Surely there will be a handfull of people spending more time than say a week if the contest runs for 2 months, but that's ok from my pov. I mean... if you manage to write an almost sellable game (exchanging content will be likely necessary) it's a good thing. Oh and having more quality java games is a good thing for the whole community, because they can ramp up the number of installed java vms on gamer PCs.

[CaptainJester]

Hmm, 2 months is a bit too long I think. Anywhere between 2 weeks to a month is reasonable. Anything over a month and you've got to start actually thinking about stuff rather than just hacking it as you go.

Given how much people get done in the LD48h compos, 2 weeks should be plenty. So a month gives a chance for multiple entries.

I disagree.  I don't get to spend as much time as I would like on the computer, so 2 months gives me enough time to do something that works.

[nonnus29]

I agree with the prohibition on p200; I won't install 1.5 to play any game.

I agree with C.J.  I think a limit on time is too unrealistic; 2 weeks too short, 1 month okay, 2 months better.  The only reason I completed an entry for the 4k is because we did have such a long time.  Most people do have alot going on in real life (tm).

Also, although no ones mentioned it, I think its very important to NOT impose a theme or content rules like ludumdare.  It should be a freeform contest, any game goes.  I hate rules in these things.

[princec]

Exactly. 16KB is your limit, you've got plenty of time and all the imagination you can muster to fill that 16KB with interesting stuff.

Cas

[woogley]

Exactly. 16KB is your limit, you've got plenty of time and all the imagination you can muster to fill that 16KB with interesting stuff.

yes, this is what I was thinking it should be the whole time. 16K should basically be a 4K except there are target libraries (that you wouldn't have to package with your game) that the users would have installed on the machine. Or, if you prefer, put whatever you want in the 16K - just use the standard java libraries. It's 4K squared, literally! And on that same note it should have at least the same amount of time that the 4K had.. seeing as there's 4 times the space to fill.

[mlk]

2 months is just right.

You should wait until the dust has really setted before running a second compo, I'd say starts beinging of June, ends end of July. Esp if we plan to hold it as a yearly event.

I like the themes idea.

[Markus_Persson]

It's 4K squared, literally!

Surely 4K squared is 16K², not 16K?

[jbanes]

Surely 4K squared is 16K², not 16K?

Depends on whether he meant (4K)², 4K², or 4²K. Each one works out as follows:

(4K)² = (4 * 4) * (K * K) = 16K²
4K² = 4 * (K * K)
4²K = (4 * 4) * K = 16K

Substituting the test value 1024 for K (duh) we can check our work as follows:

16K² = 16 * (1024 * 1024) = 16777216
(4K)² = (4 * 1024) * (4 * 1024) = 16777216

4K² = 4 * (1024 * 1024) = 4194304

4²K = (4 * 4) * 1024 = 16384
16K = 16 * 1024 = 16384

With those numbers in mind, I think woogley meant the later definition. ;-)

[Markus_Persson]

To me, saying "4k squared!" is the same as saying "four meters squared!"

4k squared == 16k^2, just as 4m squared == 16m^2
The problem is that there's no such thing as a k^2 unit.

We could define "1k == 1024 bytes", but then we just get a result of 16777216bytes^2 instead.. and there's no such thing as a byte^2.

Defining a byte ("1 byte == 8 bits!!") won't help either, as then we get 1073741824 bits^2.. and I'll be damned if I understand how a bit^2 works.

[jbanes]

To me, saying "4k squared!" is the same as saying "four meters squared!"

Effectively, yes they are the same thing. :-)

4k squared == 16k^2, just as 4m squared == 16m^2
The problem is that there's no such thing as a k^2 unit.

There isn't? Why not? Bits can be lined up in two dimensions. For example, banked memory is two dimensional in nature. Two dimensional arrays are logical bytes^2 or kb^2.

Defining a byte ("1 byte == 8 bits!!") won't help either, as then we get 1073741824 bits^2.. and I'll be damned if I understand how a bit^2 works.

Ah, quite simple! Here's 8bits^2:

1   2   3   4   5   6   7   8

1 1 1 0 1 0 1 0 0 1 1 0 1 0 1 0 1 0 1 0 1 1 1 0 1 1 0 1 1 0 1 0 1 1 1 0 0 0 1 0 1 0 1 1 1 0 0 0 0 1 0 0 1 1 1 0 1 1 1 0 1 0 1 0

I think most of your confusion comes from confusing mathematical squaring with spatial dimensions. Mathematical squaring is exactly that. A number multiplied by itself. The only difference in spatial terms is that those quantities tend to be parallel instead of additive.

[Markus_Persson]

Ah, of course!

But isn't that bit grid you posted 64 bits^2 (or 1 byte^2)?

[nonnus29]

2 months is just right.

You should wait until the dust has really setted before running a second compo, I'd say starts beinging of June, ends end of July. Esp if we plan to hold it as a yearly event.

I like the themes idea.

!!!NO THEMES!!!

Personally I will boycott any contest that enforces a theme.  The only them should be 'anything goes in 4 or 16k' as princec said earlier:

Exactly. 16KB is your limit, you've got plenty of time and all the imagination you can muster to fill that 16KB with interesting stuff.

I can't state this strongly enough.  Is it a game design competition or a hackfest?  I mean look at the variety of polished complete games in the j4k.  Now imagine if there would've been a rule like ludum dare 'there must be zombies and the gameplay element must be 'set it up and let it run' '.  If I'm going to spend 2 months making a game its got to be something that interests me.  Besides, its hard enough fitting anygame at all in 4/16k.

[mlk]

I still like the theme idea, maybe a 3rd contest...

[jojoh]

I can't state this strongly enough.  Is it a game design competition or a hackfest?

The first game I wrote for the 4K compo I spent more time fitting it in 4K than what I spent on designing the game! I therefore think that a different approach to creating a level playingfield than mem limit (possibly use theme instead) would be great, and make it LESS of a hackfest! Plus I wouldn't have to write so horribly uggly code That said, a theme that creates a level playingfield and promotes creativity and gamedesign might be very difficult to come up with.
Not sure what you mean with 'set it up and let it run'

16K > 4K but it is still a mem limit and to improve my game within the rules, I would have to do some hacking and ugglyfying my code.