The bug has disappeared, but the models still look a bit strange.
Oh jeez, not I’m really confused 
Nice to see the geometry bug has gone away but now the two models seem to have swapped skins on Linux… I guess it could be an oddity with appearances on Linux but I guess its more likely my code somewhere…
Thanks for testing it again!
Kev
This is hosed on OS X 10.3.1.
I waited a bit and I saw one static frame with an alien, brick background and “FPS: 1”
I resized the window and the frame cleared and never came back.
Do you get anything on the Java Console? Sounds like an uncaught exception.
Kev
Working fine now on Suse9.0 + NVidia. cca 200 FPS when resized fullscreen. Minor hickups when moving lef/right, every 6/7 steps or so. (I don’t have a windows box to check if that happes there too). Pretty impressive, actually!
The hickups might be GC related, haven’t really optimised (or considered) this yet.
Wahay! A positive result on a linux box… currently in the process of getting gentoo installed on the GF4 box upstairs…
Kev
java.lang.Error: Cannot load image resources/textures/stones.gif
at com.xith3d.loaders.texture.TextureLoader.loadImageFast(TextureLoader.java:293)
at com.xith3d.loaders.texture.TextureLoader.loadTexture(TextureLoader.java:313)
at com.xith3d.loaders.texture.TextureLoader.loadTexture(TextureLoader.java:348)
at com.xith3d.loaders.texture.TextureLoader.getTexture(TextureLoader.java:631)
at org.newdawn.martian.client.AppearanceStore.getAppInstance(AppearanceStore.java:32)
at org.newdawn.martian.client.AppearanceStore.getApp(AppearanceStore.java:24)
at org.newdawn.martian.client.map.Block3D.createSimpleShape(Block3D.java:161)
at org.newdawn.martian.client.map.Block3D.buildForeground(Block3D.java:143)
at org.newdawn.martian.client.map.Block3D.<init>(Block3D.java:32)
at org.newdawn.martian.client.map.World3D.<init>(World3D.java:48)
at org.newdawn.martian.client.Game.<init>(Game.java:47)
at org.newdawn.martian.client.Game.main(Game.java:120)
at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
at sun.reflect.NativeMethodAccessorImpl.invoke(Unknown Source)
at sun.reflect.DelegatingMethodAccessorImpl.invoke(Unknown Source)
at java.lang.reflect.Method.invoke(Unknown Source)
at com.sun.javaws.Launcher.executeApplication(Unknown Source)
at com.sun.javaws.Launcher.executeMainClass(Unknown Source)
at com.sun.javaws.Launcher.continueLaunch(Unknown Source)
at com.sun.javaws.Launcher.handleApplicationDesc(Unknown Source)
at com.sun.javaws.Launcher.handleLaunchFile(Unknown Source)
at com.sun.javaws.Launcher.run(Unknown Source)
at java.lang.Thread.run(Unknown Source)Weird, what platform is that Dude?
Kev
EDIT: For anyone thats interested I’ve put up some screenshots of what its meant to look like 
[quote]java.lang.Error: Cannot load image resources/textures/stones.gif
[/quote]
I receive this error a few times, too. I don’t know why it runs sometimes, sometimes it doesn’t.
Well, I think Webstart doesn’t like me (and vice versa).
Maybe it would be an option to just upload the JAR of your game? JARs are clean, fast and work (here for me) all time. I mean mainly during your development cycle and for us developers here who have installed up to date versions of Jogl and Xith3d.
PS: Apart from that the game demo is pretty funny. Good work, Kev! 
Yeah, I’ll get a download site arranged pretty soon. I just like web start cause it guaratees you’ve always got the latest version of everything.
As to having up to date Xith/Jogl etc… that can be a problem, again with web start I can assure you’re running against the version I am
Which platform are you guys seeing that error on?
Kev
quote As to having up to date Xith/Jogl etc… that can be a problem, again with web start I can assure you’re running against the version I am
[/quote]
But it won’t run here. I’ve removed all the Xith JARs from my JRE/lib/ext directory but still I got that stone.gif missing error.
[quote]Which platform are you guys seeing that error on?
[/quote]
Win2000, Java JRE 1.4.2_02-b03.
Hi Kev,
the game is running fine now. You didn’t actually change anything since yesterday, did you? The appearances aren’t swapped anymore and I can’t see any other mistake. Good work! ;D
Jens
Ok, I have a linux box! (with considerable help from Endolf). I’ve also made downloads available from the website instead of just webstart.
On linux, it seems to work flawlessly, and a sight faster too…
http://www.newdawnsoftware.com/martian
For screenies, downloads and webstart…
Kev
you can add to the readme that you start it the same way on Mac as on Windows.
Worked fine on Mac OS X 10.3.1 with a version of Java I’m not allowed to talk about here.
Superb news! and thanks for the readme info. I’ve also found the bug that makes it hang on the first frame that you were experiencing with the web start version.
However, I’m in the process of writing a simple updater that will update the installed versions from the webstart stuff so either way it should be a winner!
Thanks again,
EDIT: The updater is now posted on the website, if you could give it a go on MacOS I’d be most appreciative.
Kev
Looks like you forgot to sign something… I tried the Web Start version on the Mac and got this:
An error occurred while launching/running the application.
Title: Martian Madness
Vendor: Kev Glass - New Dawn Software
Category: Security Error
Unsigned application requesting unrestricted access to system
Unsigned resource: http://www.newdawnsoftware.com/martian/jnlp/martian.jar
I should also mention that you can just double-click martian.jar… unlike Windows that should reliably launch a jar file. (On Mac you can also double click a class file to run it’s main method.)
I did notice that there are frequent glitches in the animation. I’m going to try a few command line options (verbosegc etc…) to see if it can be smoothed out.
How useful are you!
Yep, copied over the wrong jar, fixed now
Useful info, maybe worth getting some more stuff about deployment on the Wiki. Its a tricky area now.
Yeah, I don’t see those anymore on Windows, but seems depend on the amount of memory available. I guess its GC related but I’ve not got round to worrying about that sort of thing yet. If you find anything out, it’d be really useful.
Cheers,
Kev
PS. How much is a Mac that I could test 3D stuff on now?
I did a verbosegc but didn’t see a correlation to the pauses. No FULL GC happened during the game, and the minor GCs took about 3 to 6 milliseconds.
I tried with -Xcomp but that crashed with a stack overflow and I didn’t bother to look up the args to set the stack higher.
I’m going to try with -XX:+PrintCompilation to see if it is the JITC that is causing the glitches as methods get compiled. I actually don’t think that would be the case either - methods usually compile in under a few milliseconds.
I suspect it actually related to managing the visible bits of the scene graph or something similar… but obviously I’m just guessing until there is real profiling info to go on.
Results in… nope no method compilation is causing the glitches. (FPS indicator shows between 110 and 200 on my TiBook 1GHz)
Hey is that Chicken Boo from Animaniacs on that platform?
http://www.geocities.com/RainForest/3516/theme.wav
As for a Mac to try stuff on you might be able to get a used one now that the G5’s are available and the new powerbooks are out. The new iBook looks to be good value… but you still end up paying a bit of a premium for a Mac… Worth it in my opinion, but buying used to fit your budget might make the most sense.
The low end of the new iBook is $1100 USD. A new eMac goes for $800, but I’m not sure if it has the power you would want in the graphics department
http://www.apple.com/hardware/ Street price might be a little cheaper…
Hmm
Having helped get your linux box running, I thought I ought to try it on mine :), I get the usual demp telling me what extensions are available etc, followed by java.lang.stackoverflowerror. The first time I tried it I got the all the static geometry up and visible, the last few times I get the same error, but don’t get anything visible apart from the main app frame. I tried the xith demos and they seem to be ok, no idea whats going on, but thought I would report
Endolf
Hi
Kev pointed out the ‘vertex something or other bug under linux’ and I checked, with -DXITH3D_USER_VERTEX_BUFFER_CACHING=false it all works, well, the install one does.
Endolf