Project OGLAClient O2010P1 update (OGLA 110130)

[Linguofreak]

Which dlls were that?
Are you sure you have the right fonts in the windows directory?

Intermittedness indicate Wine-related issues, or wine-exposed bugs in OGLA.

GDIPLUS.DLL and texgen.dll, IIRC. I have no idea whether they actually helped, or whether it was just intermittentness.

Unless the required fonts have changed since OGLA for P0, I'm pretty sure the fonts are good.

Wine-related issues or wine-exposed bugs in OGLA are probably a safe assumption for *any* bug I report, unless I specifically say I encountered it on our desktop (XP).

 

[Artlav]

GDIPLUS.DLL and texgen.dll, IIRC. I have no idea whether they actually helped, or whether it was just intermittentness.

texgen.dll is definitely unrelated, as it is a texture generation lib for OGalaxy.
GDIPLUS.dll could be related in some indirect way.

 

[Carmen A]

Definitely impressive results. Thanks for all the hard work :3

 

[Axel]

Ok i updated my catalyst now to the actual 11.01. And OGLA runs! I have 85 fps with high settings. But i see a very slow kind of "microtexture" spawning/building overall, around the scenario place. Its like the microtexture in orulex, but mutch more slower. Is there a possibility to speed up or deactivate the microtextures? When i zoom out of the area the switch from OGLA-texture/terrain to earth-texture is very fast and unsmooth, the brightness switches from dark to light too very fast. But OGLA runs not every time, only at every second loading, without changing of options, thats crazy! I cant read the readings at the loading screen, when i get a CTD, with the message "error at initializing OpenGL" and "Runtime error" and "ram access error". I registered that my PC is slower and slower after every OGLA start, maybe it takes alot memory? All windows are opening slower ect. Maybe it steals memory? I use no V-Ram, because i have 4 Gb real ram.

A very nice idea with OGLA, but needs still alot of work and bugfixing. And a collision detection later, at the moment i prefer normal orbiter with orulex. A question, why do you use OpenGL instead DirectX? DirectX can do all your features and more, and that faster! OpenGL is only at the level from DirectX8. Orfcourse the normal orbiter uses directx7 from year 2000, thats ... no words. But there is a guy who writes a DX9 driver for it, you could do it too or maybe work with him together?

Here the log, when it crashes to desktop.

01.02.2011-15:51:45:[START     ]:-------------------------------------------------------
01.02.2011-15:51:45:[INIT      ]:OGLA v110130 (GLGR)
01.02.2011-15:51:53:[GPU INFO  ]:Name=
01.02.2011-15:51:53:[GPU INFO  ]:Description=
01.02.2011-15:51:53:[GPU INFO  ]:AdapterCompatibility=
01.02.2011-15:51:53:[GPU INFO  ]:AdapterRAM=0 Mb
01.02.2011-15:51:53:[GPU INFO  ]:VideoModeDescription=
01.02.2011-15:51:53:[GPU INFO  ]:VideoProcessor=
01.02.2011-15:51:53:[GPU INFO  ]:DriverVersion=
01.02.2011-15:51:53:[INIT      ]:GPU supports OpenGL 1.2
01.02.2011-15:51:53:[INIT      ]:GPU supports OpenGL 1.4
01.02.2011-15:51:53:[INIT      ]:GPU supports OpenGL 2.0
01.02.2011-15:51:53:[INIT      ]:glgruva=1, usevbo=1, vboav=1, glgr_stensh_aupd=0, gvsync=0, gl_comp_sup=1, gl_shm4=0
01.02.2011-15:51:54:[OGLADBG   ]:OGLA 110130 Debug.
Haze program 0:
Vertex shader was successfully compiled to run on hardware.

Fragment shader was successfully compiled to run on hardware.

Vertex shader(s) linked, fragment shader(s) linked.

Ring program:
Vertex shader was successfully compiled to run on hardware.

Fragment shader was successfully compiled to run on hardware.

Vertex shader(s) linked, fragment shader(s) linked.

Post program:
Vertex shader was successfully compiled to run on hardware.

Fragment shader was successfully compiled to run on hardware.

Vertex shader(s) linked, fragment shader(s) linked.

Haze program 0:
Vertex shader was successfully compiled to run on hardware.

Fragment shader was successfully compiled to run on hardware.

Vertex shader(s) linked, fragment shader(s) linked.

Ring program:
Vertex shader was successfully compiled to run on hardware.

Fragment shader was successfully compiled to run on hardware.

Vertex shader(s) linked, fragment shader(s) linked.

Post program:
Vertex shader was successfully compiled to run on hardware.

Fragment shader was successfully compiled to run on hardware.

Vertex shader(s) linked, fragment shader(s) linked.

01.02.2011-15:51:58:[PLNTR     ]:Loading Textures/Saturn_ring.tex OK
01.02.2011-15:51:58:[PLNTR     ]:Loading Textures/Uranus_ring.tex OK
01.02.2011-15:51:58:[Del_Q     ]:Undefined error
01.02.2011-15:51:59:[Graph     ]:Error: Error in proc_smob (i=33)
01.02.2011-15:51:59:[OGLA      ]:Error: Error in oglareupdatevals

 

[Artlav]

I registered that my PC is slower and slower after every OGLA start

Sounds familiar.
Go to launchpad -> extra -> debugging options -> Orbiter shutdown options.
If it's not set to terminate Orbiter process then all hell could break loose every some time you restart the game.

Is there a possibility to speed up or deactivate the microtextures?

Press F7 in game. That opens a menu, next key press is considered a command in that menu. What you're looking for is 5 - terrain.

And a collision detection later

Collision detection i'm capable of doing is incompatible with Orbiter.

DirectX can do all your features and more, and that faster! OpenGL is only at the level from DirectX8.

These facts are obsolete by just under a decade, please research them again.

Also, OpenGL is more streamlined and intuitive than DirectX, at least from my point of view.

Other reasons - i tend to prefer open standards that can be supported regardless of companies collapsing, merging or changing their mind.

Nice side effect is platform-independence.

 

[Linguofreak]

A very nice idea with OGLA, but needs still alot of work and bugfixing. And a collision detection later, at the moment i prefer normal orbiter with orulex. A question, why do you use OpenGL instead DirectX? DirectX can do all your features and more, and that faster! OpenGL is only at the level from DirectX8. Orfcourse the normal orbiter uses directx7 from year 2000, thats ... no words. But there is a guy who writes a DX9 driver for it, you could do it too or maybe work with him together?

Have you tried the previous version of OGLA with Orbiter2010 P0? That was working much better for me.

 

[jimblah]

Looking at virtual machines too.

Any version of wine you want to test can be found here:
http://www.ibiblio.org/pub/linux/system/emulators/wine/
My best results so far have have been with 1.3.9
Tried Virtualbox today but without oglaclient, it was a no go. See http://www.dedoimedo.com/computers/virtualization-3d-support-virtualbox.html Thought I might look at vm's too.

 

[Linguofreak]

I've determined that the crash I'm getting has to do with the 2D Delta Glider panel. The generic cross-spacecraft hud panel seems to be fine, and the Shuttle-A 2D panel is glitchy but doesn't crash. 3D cockpits seem to work. So it seems that something is wrong with 2D panel handling.

 

[n72.75]

I've determined that the crash I'm getting has to do with the 2D Delta Glider panel. The generic cross-spacecraft hud panel seems to be fine, and the Shuttle-A 2D panel is glitchy but doesn't crash. 3D cockpits seem to work. So it seems that something is wrong with 2D panel handling.

I think I've been experiencing the same thing.

 

[jimblah]

Using virtual box with ogla in xp and getting good functionality with orbiter. Only 7-11 fps though as opposed to better frame rates with wine. The f7 panel is cool. The main thing is functionality: UMmu,DGIV, OMP,solid sound. The tutorials in the cockpit are awesome. Rendev with ISS right now.

 

[Linguofreak]

I think I've been experiencing the same thing.

And this on Windows? That would mean we've found a bug that's broader than just Wine.

Machine specs?

I've got

Core 2 Duo T6500 @ 2.1 GHz
3 gigs RAM (reported in Ubuntu as 2.9)
GeForce G 105M with 512 megs video RAM
Wine 1.3.11 on Ubuntu 10.04

 

[Rtyh-12]

So... any idea on when this could be considered "playable"? It looks cool but it seems that there are still some bugs, right?

 

[Linguofreak]

So... any idea on when this could be considered "playable"? It looks cool but it seems that there are still some bugs, right?

Anywhere from months ago to months from now, depending which version you're using and the quirks of your system.

With the previous version of OGLA on Orbiter2010 P0, things were very much playable for me. The current version on P1 is giving me major problems.

Also note that I'm running this under Wine on Linux, so you may very well see me report bugs that don't apply to many other people.

 

[jimblah]

I just uploaded some screenshots into the unedited images folder here on the forum. Best results so far. Wine's pro's: better frame rates, poor functionality. Virtualbox has really good functionality but lower frame rates. Native XP does it all. These shots were taken with Virtualbox 3.5; what was in the repository. 4.0 claims better Opengl support. I've installed a 4.0 binary but I'm compiling as well.
Latest Orbiter, latest Oglaclient, soon to be latest VirtualBox.

3.5Ghz from 3.2GhzP4, P4P800E-Deluxe, 2GB RAM, Nvidia 8400 GS. I'm installing another 2GB of RAM tomorrow. I might bump the processor up a little. As it stands I can already run Orbiter as long as I want now in linux but it's still a thought exercise. For the "real action" I still need native XP.

 

[n72.75]

And this on Windows? That would mean we've found a bug that's broader than just Wine.

Machine specs?

I've got

Core 2 Duo T6500 @ 2.1 GHz
3 gigs RAM (reported in Ubuntu as 2.9)
GeForce G 105M with 512 megs video RAM
Wine 1.3.11 on Ubuntu 10.04

P8400 @ 2.7GHz
4096MB RAM
2x ATI 3870s with 512MB ram each.
Running Windows 6.0.6002 64bit.

 

[bpops]

Amazing as always, Artlav. I was wondering if there's a way to shrink down the size of the HUD elements (e.g. pitch ladder, compass...).

I tried changing the panel size to something different, but anything other than 1 produces a CTD on startup.

I'm running at 3840 x 1024, so by default the HUD is huge (see attached).

 

[Columbia42]

Artifact scenario works but crashes when I switch to cockpit view in the DG.

Ogla log:

2/6/2011-6:14:58 PM:[START     ]:-------------------------------------------------------
2/6/2011-6:14:58 PM:[INIT      ]:OGLA v110130 (GLGR)
2/6/2011-6:15:02 PM:[GPU INFO  ]:Name=RADEON X300 SE 128MB HyperMemory Secondary
2/6/2011-6:15:02 PM:[GPU INFO  ]:Description=RADEON X300 SE 128MB HyperMemory Secondary
2/6/2011-6:15:02 PM:[GPU INFO  ]:AdapterCompatibility=ATI Technologies Inc.
2/6/2011-6:15:02 PM:[GPU INFO  ]:AdapterRAM=128 Mb
2/6/2011-6:15:03 PM:[GPU INFO  ]:VideoModeDescription=1024 x 768 x 4294967296 colors
2/6/2011-6:15:03 PM:[GPU INFO  ]:VideoProcessor=RADEON X300 SE (0x5B60)
2/6/2011-6:15:03 PM:[GPU INFO  ]:DriverVersion=6.14.10.6561
2/6/2011-6:15:03 PM:[INIT      ]:GPU supports OpenGL 1.2
2/6/2011-6:15:03 PM:[INIT      ]:GPU supports OpenGL 1.4
2/6/2011-6:15:03 PM:[INIT      ]:GPU supports OpenGL 2.0
2/6/2011-6:15:03 PM:[INIT      ]:glgruva=1, usevbo=1, vboav=1, glgr_stensh_aupd=0, gvsync=0, gl_comp_sup=1, gl_shm4=0
2/6/2011-6:15:03 PM:[OGLADBG   ]:OGLA 110130 Debug.
2/6/2011-6:15:07 PM:[PLNTR     ]:Loading Textures/Saturn_ring.tex OK
2/6/2011-6:15:07 PM:[PLNTR     ]:Loading Textures/Uranus_ring.tex OK
2/6/2011-6:15:08 PM:[PLNT      ]:Loading Textures/Earth.tex OK
2/6/2011-6:15:08 PM:[PLNT      ]:Loading Textures/Earth_cloud.tex OK
2/6/2011-6:15:08 PM:[PLNT      ]:Loading Textures/Earth_lmask.tex OK
2/6/2011-6:15:08 PM:[PLNT      ]:Loading Textures/Moon.tex OK

When switching to cockpit view with the test_pill vessel the HUD is fine but MFDs are all gibberish.

Cubic ogla test scenario crashes on startup.

Log:

2/6/2011-6:22:09 PM:[START     ]:-------------------------------------------------------
2/6/2011-6:22:09 PM:[INIT      ]:OGLA v110130 (GLGR)
2/6/2011-6:22:11 PM:[GPU INFO  ]:Name=RADEON X300 SE 128MB HyperMemory Secondary
2/6/2011-6:22:12 PM:[GPU INFO  ]:Description=RADEON X300 SE 128MB HyperMemory Secondary
2/6/2011-6:22:12 PM:[GPU INFO  ]:AdapterCompatibility=ATI Technologies Inc.
2/6/2011-6:22:12 PM:[GPU INFO  ]:AdapterRAM=128 Mb
2/6/2011-6:22:12 PM:[GPU INFO  ]:VideoModeDescription=1024 x 768 x 4294967296 colors
2/6/2011-6:22:12 PM:[GPU INFO  ]:VideoProcessor=RADEON X300 SE (0x5B60)
2/6/2011-6:22:12 PM:[GPU INFO  ]:DriverVersion=6.14.10.6561
2/6/2011-6:22:12 PM:[INIT      ]:GPU supports OpenGL 1.2
2/6/2011-6:22:12 PM:[INIT      ]:GPU supports OpenGL 1.4
2/6/2011-6:22:12 PM:[INIT      ]:GPU supports OpenGL 2.0
2/6/2011-6:22:12 PM:[INIT      ]:glgruva=1, usevbo=1, vboav=1, glgr_stensh_aupd=0, gvsync=0, gl_comp_sup=1, gl_shm4=0
2/6/2011-6:22:12 PM:[OGLADBG   ]:OGLA 110130 Debug.
2/6/2011-6:22:16 PM:[PLNTR     ]:Loading Textures/Saturn_ring.tex OK
2/6/2011-6:22:16 PM:[PLNTR     ]:Loading Textures/Uranus_ring.tex OK

I have minimum graphics features selected and am running ATI RADEON X300 on Windows XP.

 

[Linguofreak]

I've seen one other thread and one bug report dealing with a "CTD on switch to 2D panel" issue that seems to be identical to the one me, n72.75, and Columbia42 have commented on on this thread.

Both were made by users running the built-in client on Windows. It seems that this issue has nothing whatsoever to do with OGLAClient.

EDIT:

Actually, from this thread it seems that the problem is that the DG in the 110130 Vessel Demo Package is compiled against P0 instead of P1. Could this be corrected?

 

[bpops]

Linguofreak: You are correct, this issue has nothing to do with OGLAClient. I originally thought it did, which is why I posted it in that other thread.

I am at work now, but will try to update this DLL as soon as possible and check if that works.

Thanks.

 

[jimblah]

I bring good news...

:theysaid: It took WineHQ 7 days to validate my most recent test submission for Orbiter. Orbiter has its' first gold rating at Wine Appd. I'm thrilled. This is reason to celebrate.

http://appdb.winehq.org/objectManager.php?sClass=version&iId=22675