LWJGL 0.9 released!

[elias]

I have:

/usr/lib/libstdc++.so.2.7.2
/usr/lib/libstdc++.so.2.7.2.8
/usr/lib/libstdc++.so.2.8
/usr/lib/libstdc++.so.2.8.0
/usr/lib/libstdc++.so.2.9
/usr/lib/libstdc++.so.2.9.0
/usr/lib/libstdc++.so.5
/usr/lib/libstdc++.so.5.0.5

Seems like you have the same... What nvidia driver version do you use? I've got 5336

- elikas

[AndersD]

Yep, me too.

[erikd]

Thanks to tomb, I checked in a fix for scaling images up.
It's a temporary fix; it doesn't do any filtering yet. I'll see if I can add that too. I think I can still get inspired by the mesa code just as long as don't copy paste their code, but that code is of course not the only way to do filtered up-scaling.

Apologies for the delay. For some reason I can only access CVS for LWJGL at work. I get some authentication error messages at home when I try to connect with LWJGL's CVS. I can still connect to another project at home though. Dunno why...  :roll:

[erikd]

BTW, parts of the mipmap code still has the mesa code as a starting point but it's mostly a rewrite now. Some variable names and such are still the same though. I don't think this should be a problem or is it?...

[elias]

what are the error messages?

- elias

[Orangy Tang]

I'm seeing some mouse button oddities when using the latest build. I've got a windowed display alongside a couple of Swing windows and need to switch between the two of them. I'm using the native cursor and it was working ok (although the display window did cause some focus oddities when switching windows).

When the app starts all the mouse buttons appear 'jammed down', only when one is actually pressed does it unjam things and they all work correctly. I seem to get the same thing when I switch the focus away and back to the LWJGL window.

The mouse test app doesn't have this problem, but that actually captures the mouse and doesn't use the native cursor. This all did work when I was using a month old build, and nothing else has really changed.

Mouse is created and then I let window.update poll it for me. No buffering or anything fancy either. Anyone any ideas?

[elias]

Did you try the Mouse test when commenting out the Mouse.setGrabbed(true) call?

- elias

[Orangy Tang]

Actually I got that wrong, the mouse test doesn't grab the mouse, it just leaves it alone on the regular native cursor.

Either way, I tried it again with some text output and it seems that clicking outside to make it loose focus triggers the weird behaviour: all the buttons are magically held down and stay that way until it regains focus.

Regaining focus seems to fix things on the test app, regardless of whether I click back or alt-tab back.

[elias]

I posted a fix to the win32 tree (polled mouse fields were filled even when the mouse is lost). Are you able to compile from CVS?

- elias

[Orangy Tang]

I can compile from CVS, but Eclipse insists none of the files have changed (at least, no mouse related ones have). What files should I be looking for?

[elias]

Yoy're out of luck - I changed the native side of win32...

- elias

[Orangy Tang]

Native is not a problem, I've got VS set up so I can recompile the Win32 dll if need be. I'm assuming you've changed src/native/win32/*mouse.cpp, but I don't see any later revisions that the one from Matzon on the 12th of april. Wonder if Sourceforge is just being slow...

[elias]

Ah I seem to remember that anonymous CVS is a day or so behind developer CVS...

- elias

[Orangy Tang]

Cheers, that seems to have fixed things

[spasi]

Nvidia released some new extensions and I just commited them to cvs. These are:

EXT_blend_equation_separate
EXT_texture_mirror_clamp
GL_NV_fragment_program2
GL_NV_vertex_program2_option
GL_NV_vertex_program3

A little request:

Could somebody please remove anything that's not used from cvs? There are many useless folders and the extgl.cpp has a lot of crap in it. A cleanup would be nice. I could do it, but I want to be sure they are not needed first.