It seems I cant download some jars when starting the webstart.

alright  

Wondering if there is still any development active of Xith?
Since we are using it for magicosm, I surely hope so.

Nice contribution!  

Look at what extra APIs are available for the device you are developing your application for. It is not part of MIDP 1.0 or 2.0 as far as I know.

The bigass robot? It is part of the images you can choose from here on JGO.

I am already using it in my 2 games and it works very well. Only downside to this library is the speed at which the fonts are rendered. But for menus or RPG/games where you dont need max fps at that moment, it works great.

Not possible with pure MIDP 1.0 tho, is it?

What is a QA team?

Looks great, works great, great fun
AMD TBird 1.4 Gforce TI4200.

You should learn not only SUNs documents but also some other (aspecially Nokia) as you will be needing them when tailoring games aspecially for devices or even to make sure they are really cross platform as the behaviour is different from device to device.

That is not part of MIDP 1 or 2.. And I doubt any extra API gives you that functionality. However I know there is something similar to what you want(called the localization API or something), but I do not remember the name.

Reading through the MIDP 1.0 API should give you a very good start. Then the source of the few examples included in the Wireless Toolkit. You can get it all from java.sun.com under J2ME.

I am looking for a distributor for my game "E-Scape" that won 4th place in the Excitera Mobile awards 2004 (www.excitera.nu). If you are interested, I can send you a demo version, just e-mail me at darcone@qiui.com.

If you want, I can send you my own implementation of vector that is much cheaper and faster than the default one. Just e-mail me at darcone@qiui.com.

Theres no ArrayList in J2ME only Vector and it´s pretty expensive... :-/

I put up a test midlet for you to try and compare on different emulators etc at http://www.qiui.com/MicroFont/java/MicroFontTest.jar & http://www.qiui.com/MicroFont/java/MicroFontTest.jad

Im learning ANT atm so I wont be able to upload a test version for a few days.

The fonts are compiled into binary files and the J2ME class reads that info and creates arrays, yes. Thus you can include many different fonts, sizes etc into the same midlet and use the same font class. Well, I intend it to be free for educational use and offer commercial licenses. A free version should be up by tonight.

I dont understand what you mean by that  :-/
I have not yet decided if I am going to release the font reader(the class that prints the fonts) under GPL, but the compiler I will probably not for some time.

Ive been working on this the last few days. Check it out and tell me what you think.

http://www.qiui.com/MicroFont/

First of all you should be aware of the fact that MIDP 1.0 doesnt give you access to TCP by default, only HTTP and UDP iirc.

Ok, I take that back

Indeed I always find it useful to see other peoples code that they know in practice. As we all know, making things work for J2ME isnt always as easy as it seems at first.

I ran a test comparing two methods on a J2ME emulator:

Method 1:



Method 2:



After a few hundred tests (didnt really believe it myself after just a few tests) I got this result on 10000 elements in a int[]:

1: 150 ms
2: 80 ms

So apparently I was right about the ugly method 2 being faster. Now the question remains if it is faster to do a "i != 0" comparison rather than doing a "i < b" comparison, my argument was that doing a "!= 0" would only need one memory reference, while the second would need two. But erik said it wouldnt matter since of the pipeline architecture and since I am not an expert when it comes to hardware perhaps someone has somethint to say about this?

Hmm sounds like something you can do with a array and some magic
*wishes for structs in java*

Im pretty sure (at least for now) that the most efficient way to write J2ME apps is from the bottom up without using any special libs. But keep up the good work, perhaps some day memory and performance restrictions will make your package a great tool.

What is a sprite class?  

Regarding that J2ME for loop - first, I believe it would be faster to do a:



Thats what I have read on Sony ericsson slides. Another interesting thing I read was that due to the way exceptions are being handled on a J2ME platform, this should be a *bit* faster too:



I may be wrong, but I may be right And again, this (aspecially the last one) applies to J2ME, the first one to J2SE but is more ugly than effective.

If the GPU is not the bottleneck but the CPU is, then its not really JOGLs fault, is it? I mean, if you cant keep the GPU busy enough (by comparison) that means you shouldnt worry about JOGL performance?

Yeah, and make the highscore only list last few days entries so that cheaters will have to do it every week (in which case you can identify and hopefully ban them)