New Release OGLA OpenGL Client 080919

Artlav · Sep 16, 2008

Current version: 080919, downloadable in post 18

Downloadable here:
orbides.1gb.ru/orbf/oglaclient-080916_beta-080916.zip (1.0 Mb)
For beta 080910 with 080916 update.

(Edit: Hmm. Since the beta is public, why keep it closed?
Update 080916 for beta is mirrored here: http://orbides.1gb.ru/orbf/orbiter_080916.zip (1.5 Mb)
Where the beta is for public access - you find it

.)

After some long reworking and despaghetting, pinpointing and fixing old and new bugs in core and client alike, OGLAClient works fine again.

Now, most of the scenarios should work, not just a set of Shuttle-A_and_etc_clean ones, thanks to Martins fixing major core bugs and leftovers.
Known incompatibilities - Dragonfly and Hubble from default set.

Some features are not fully functional yet:
*Lv9+ texture format was changed subtly, and not up to speed yet.
*Terrain support (Orulex) is working, but known to crash out of the blue, as Orulex is in the middle of a rewrite, so it is disabled by default (see Visual effects tab to enable).
*Automatic shadow selector is not sane yet. There are 5 shadow modes based on three algorithms, one of which is new and fast in hi-end scenes, but slower than the olders in small scenes, so switching helps a lot.
*GDI emulation works fine, allowing MFD's and HUD's to work at 40% to 60% FPS penalty, but old panel code is not supported.

You can set different graphical configs in Visual Config tab new button, or use cfg file.

All in all, this release is to show that the progress is here.
Since there was a while since i released a stabilized OGLAClient, performance and visuals feedback is welcome.
I want to know, if it works anywhere else now.

Some other add-ons compatibility testing may make sense at this stage, especially non-panelled ones.

Some enumerable changes are:
*Reimplemented shadows - mode 5 is 3x faster than mode 4 at same quality
*All animation support
*Mesh loading shifts and fixes support (SS Atlantis looks fine now)
*General speed-ups
*Several design improvements
*Major refurbishing
*Visual settings config tab

What is reasonable to expect:
080919:
Orbital:

KSC Update:

Cruising above Florida:

Close-up on Shuttle:

080916:
Full-wide, with GDI emulation on:

Similar, GDI emulation off:

For the Shuttle fans:

And default scenario:

tblaxland · Sep 17, 2008

I've been so used to Orbiter's lack of self-shadowing that I'd forgotten what a difference it makes. Nice work, Artlav.

DaveS · Sep 17, 2008

Hmmm, not exactly working nicely here. Using latest Orbiter beta but all I get is black meshes and the following in the Ogla.log:

Code:

2008-09-17 22:56:22|    INIT| OGLA, GLGR, VBO 080916
2008-09-17 22:57:24|    INIT| OGLA, GLGR, VBO 080916
2008-09-17 22:57:31|    PLNT| Loading Textures/Earth.tex OK
2008-09-17 22:57:31|    PLNT| Loading Textures/Earth_cloud.tex OK
2008-09-17 22:57:31|    PLNT| Loading Textures/Moon.tex OK
2008-09-17 22:57:31|   PLNTR| Loading Textures/Saturn_ring.tex OK
2008-09-17 22:57:31|   PLNTR| Loading Textures/Uranus_ring.tex OK
2008-09-17 22:57:31|   Graph| Error: Error in ldmsh (i=0, gc=0)
2008-09-17 22:57:31|   Graph| Error: Error in procvess (i=3)
2008-09-17 22:57:31|   Graph| Error: Error in drawvess (i=3)
2008-09-17 22:57:31|   Graph| Error: Error in procvess (i=3)
2008-09-17 22:57:31|   Graph| Error: Error in drawvess (i=3)
2008-09-17 22:57:31|   Graph| Error: Error in procvess (i=3)
2008-09-17 22:57:31|   Graph| Error: Error in drawvess (i=3)
2008-09-17 22:57:31|   Graph| Error: Error in procvess (i=3)
2008-09-17 22:57:31|   Graph| Error: Error in drawvess (i=3)
2008-09-17 22:57:31|   Graph| Error: Error in procvess (i=3)
2008-09-17 22:57:31|   Graph| Error: Error in drawvess (i=3)
2008-09-17 22:57:31|   Graph| Error: Error in procvess (i=3)
2008-09-17 22:57:31|   Graph| Error: Error in drawvess (i=3)
2008-09-17 22:57:31|   Graph| Error: Error in procvess (i=3)
2008-09-17 22:57:31|   Graph| Error: Error in drawvess (i=3)
2008-09-17 22:57:31|   Graph| Error: Error in procvess (i=3)
2008-09-17 22:57:31|   Graph| Error: Error in drawvess (i=3)
2008-09-17 22:57:31|   Graph| Error: Error in procvess (i=3)
2008-09-17 22:57:31|   Graph| Error: Error in drawvess (i=3)
2008-09-17 22:57:31|   Graph| Error: Error in procvess (i=3)
2008-09-17 22:57:31|   Graph| Error: Error in drawvess (i=3)
2008-09-17 22:57:54|    INIT| OGLA, GLGR, VBO 080916
2008-09-17 22:58:11|    PLNT| Loading Textures/Earth.tex OK
2008-09-17 22:58:11|    PLNT| Loading Textures/Earth_cloud.tex OK
2008-09-17 22:58:11|    PLNT| Loading Textures/Moon.tex OK
2008-09-17 22:58:11|   PLNTR| Loading Textures/Saturn_ring.tex OK
2008-09-17 22:58:11|   PLNTR| Loading Textures/Uranus_ring.tex OK

.
This is with the 2008 Edition\Visual Improvements\KSC Update scenario. Any Atlantis scenario crashes with the same log messages.

Artlav · Sep 18, 2008

An update with some fixes and debug info on:
http://orbides.1gb.ru/orbf/oglaclient-080918_beta-080916.zip

DaveS said:
Using latest Orbiter beta but all I get is black meshes

Does it look something like that per chance?

What if you turn shadows off or into different mode (F7-7)?
What if you turn advanced graphics off (F7-U)?

Is there the same behavior with the update?

DaveS said:
and the following in the Ogla.log:

Code:

... 2008-09-17 22:57:31| Graph| Error: Error in ldmsh (i=0, gc=0) ...

.
This is with the 2008 EditionVisual ImprovementsKSC Update scenario. Any Atlantis scenario crashes with the same log messages.

Are there any scenarios that does work fine, or it crashes/look black in all of them?
Is there the same behavior with the update?
If the problem remains, please provide the log from after update run.

DaveS · Sep 18, 2008

Well, now the pitch black meshes are gone and replace with invisible meshes. Turning of advanced graphics cures that. Lost exhaust, contrail and surftiles though.

Contents of OLGA.log after a run with the updated package:

Code:

2008-09-17 22:56:22|    INIT| OGLA, GLGR, VBO 080916
2008-09-17 22:57:24|    INIT| OGLA, GLGR, VBO 080916
2008-09-17 22:57:31|    PLNT| Loading Textures/Earth.tex OK
2008-09-17 22:57:31|    PLNT| Loading Textures/Earth_cloud.tex OK
2008-09-17 22:57:31|    PLNT| Loading Textures/Moon.tex OK
2008-09-17 22:57:31|   PLNTR| Loading Textures/Saturn_ring.tex OK
2008-09-17 22:57:31|   PLNTR| Loading Textures/Uranus_ring.tex OK
2008-09-17 22:57:31|   Graph| Error: Error in ldmsh (i=0, gc=0)
2008-09-17 22:57:31|   Graph| Error: Error in procvess (i=3)
2008-09-17 22:57:31|   Graph| Error: Error in drawvess (i=3)
2008-09-17 22:57:31|   Graph| Error: Error in procvess (i=3)
2008-09-17 22:57:31|   Graph| Error: Error in drawvess (i=3)
2008-09-17 22:57:31|   Graph| Error: Error in procvess (i=3)
2008-09-17 22:57:31|   Graph| Error: Error in drawvess (i=3)
2008-09-17 22:57:31|   Graph| Error: Error in procvess (i=3)
2008-09-17 22:57:31|   Graph| Error: Error in drawvess (i=3)
2008-09-17 22:57:31|   Graph| Error: Error in procvess (i=3)
2008-09-17 22:57:31|   Graph| Error: Error in drawvess (i=3)
2008-09-17 22:57:31|   Graph| Error: Error in procvess (i=3)
2008-09-17 22:57:31|   Graph| Error: Error in drawvess (i=3)
2008-09-17 22:57:31|   Graph| Error: Error in procvess (i=3)
2008-09-17 22:57:31|   Graph| Error: Error in drawvess (i=3)
2008-09-17 22:57:31|   Graph| Error: Error in procvess (i=3)
2008-09-17 22:57:31|   Graph| Error: Error in drawvess (i=3)
2008-09-17 22:57:31|   Graph| Error: Error in procvess (i=3)
2008-09-17 22:57:31|   Graph| Error: Error in drawvess (i=3)
2008-09-17 22:57:31|   Graph| Error: Error in procvess (i=3)
2008-09-17 22:57:31|   Graph| Error: Error in drawvess (i=3)
2008-09-17 22:57:54|    INIT| OGLA, GLGR, VBO 080916
2008-09-17 22:58:11|    PLNT| Loading Textures/Earth.tex OK
2008-09-17 22:58:11|    PLNT| Loading Textures/Earth_cloud.tex OK
2008-09-17 22:58:11|    PLNT| Loading Textures/Moon.tex OK
2008-09-17 22:58:11|   PLNTR| Loading Textures/Saturn_ring.tex OK
2008-09-17 22:58:11|   PLNTR| Loading Textures/Uranus_ring.tex OK
2008-09-17 23:00:16|    INIT| OGLA, GLGR, VBO 080916
2008-09-17 23:05:36|    INIT| OGLA, GLGR, VBO 080916
2008-09-17 23:05:44|    INIT| OGLA, GLGR, VBO 080916
2008-09-18 11:27:54|    INIT| OGLA, GLGR, VBO 080918
2008-09-18 11:28:05| OGLADBG| OGLA 080918 Debug.
2008-09-18 11:28:05|OGLASHDR| Vessel program:
2008-09-18 11:28:05|OGLASHDR| Fragment info
-------------
(71) : warning C7548: % requires "#extension GL_EXT_gpu_shader4 : enable" before use
(70) : warning C7548: % requires "#extension GL_EXT_gpu_shader4 : enable" before use
(70) : warning C7548: % requires "#extension GL_EXT_gpu_shader4 : enable" before use
(71) : warning C7548: % requires "#extension GL_EXT_gpu_shader4 : enable" before use
(71) : warning C7548: % requires "#extension GL_EXT_gpu_shader4 : enable" before use
(70) : warning C7548: % requires "#extension GL_EXT_gpu_shader4 : enable" before use
(70) : warning C7548: % requires "#extension GL_EXT_gpu_shader4 : enable" before use
(71) : warning C7548: % requires "#extension GL_EXT_gpu_shader4 : enable" before use
2008-09-18 11:28:05|OGLASHDR| Planet program:
2008-09-18 11:28:05|OGLASHDR| Fragment info
-------------
(71) : warning C7548: % requires "#extension GL_EXT_gpu_shader4 : enable" before use
(70) : warning C7548: % requires "#extension GL_EXT_gpu_shader4 : enable" before use
(70) : warning C7548: % requires "#extension GL_EXT_gpu_shader4 : enable" before use
(71) : warning C7548: % requires "#extension GL_EXT_gpu_shader4 : enable" before use
(71) : warning C7548: % requires "#extension GL_EXT_gpu_shader4 : enable" before use
(70) : warning C7548: % requires "#extension GL_EXT_gpu_shader4 : enable" before use
(70) : warning C7548: % requires "#extension GL_EXT_gpu_shader4 : enable" before use
(71) : warning C7548: % requires "#extension GL_EXT_gpu_shader4 : enable" before use
2008-09-18 11:28:05|OGLASHDR| Ring program:
2008-09-18 11:28:05|OGLASHDR| Fragment info
-------------
(90) : warning C7548: % requires "#extension GL_EXT_gpu_shader4 : enable" before use
(90) : warning C7548: % requires "#extension GL_EXT_gpu_shader4 : enable" before use
(90) : warning C7548: % requires "#extension GL_EXT_gpu_shader4 : enable" before use
(90) : warning C7548: % requires "#extension GL_EXT_gpu_shader4 : enable" before use
2008-09-18 11:28:05|OGLASHDR| Clouds program:
2008-09-18 11:28:05|OGLASHDR| Fragment info
-------------
(100) : warning C7548: % requires "#extension GL_EXT_gpu_shader4 : enable" before use
(100) : warning C7548: % requires "#extension GL_EXT_gpu_shader4 : enable" before use
(100) : warning C7548: % requires "#extension GL_EXT_gpu_shader4 : enable" before use
(100) : warning C7548: % requires "#extension GL_EXT_gpu_shader4 : enable" before use
2008-09-18 11:28:05|OGLASHDR| Star program:
2008-09-18 11:28:05|OGLASHDR| Fragment info
-------------
(32) : warning C7548: % requires "#extension GL_EXT_gpu_shader4 : enable" before use
(32) : warning C7548: % requires "#extension GL_EXT_gpu_shader4 : enable" before use
(32) : warning C7548: % requires "#extension GL_EXT_gpu_shader4 : enable" before use
(32) : warning C7548: % requires "#extension GL_EXT_gpu_shader4 : enable" before use

Computer specs:
CPU: Intel Core2 Quad Q6600, ~2.4 GHz
System RAM: 3 GB
Video card: NVIDIA GeForce 8600 GT with latest drivers
OS: Widnows Visat SP1.

Artlav · Sep 18, 2008

DaveS said:
Turning of advanced graphics cures that.
...
(71) : warning C7548: % requires "#extension GL_EXT_gpu_shader4 : enable" before use
...

Ok, that clears the location of the problem... Or i hope so.
Can you try this build for that problem?
http://orbides.1gb.ru/orbf/oglaclient-080918_beta-080916-b.zip

What about the crash?
Does it still exists?
If yes, could you provide a log with it?

DaveS · Sep 18, 2008

The CTD with Atlantis is gone.

Artlav · Sep 18, 2008

An update:
http://orbides.1gb.ru/orbf/oglaclient-080918-c_beta-080918.zip

Updated to current beta 080918.
Cleaned up debug code from previous report.

DaveS · Sep 18, 2008

Artlav said:
An update:
http://orbides.1gb.ru/orbf/oglaclient-080918-c_beta-080918.zip

Updated to current beta 080918.
Cleaned up debug code from previous report.

Now the Orbiter launchpad CTDs. This in the Orbiter.log:

Code:

**** Orbiter.log
Build Sep 16 2008 [v.080916]
Timer precision: 6.98413e-008 sec
Found 1 joystick(s)
Module AtlantisConfig.dll [API v.071208]
Module DGConfig.dll [API v.080812]
Module OGLAClient.dll [API v.080812]
Error loading module ModulesPluginOGLAClient.dll (code 998)

Artlav · Sep 18, 2008

DaveS said:
Build Sep 16 2008 [v.080916]

Please take a closer look on the file name and the content of the previous post...

DaveS · Sep 18, 2008

Artlav said:
Please take a closer look on the file name and the content of the previous post...

Well, that fixed things! Still no base textures though.

Artlav · Sep 18, 2008

DaveS said:
Well, that fixed things! Still no base textures though.

That is good, but whether it is a new problem or the original one, could you post some screenshots of the problems?

DaveS · Sep 18, 2008

Screenshot of problem is attached to this post.

Artlav · Sep 18, 2008

Oh, that.
Not updated yet, as i said in the first post

*Lv9+ texture format was changed subtly, and not up to speed yet.
*Terrain support (Orulex) is working, but known to crash out of the blue, as Orulex is in the middle of a rewrite, so it is disabled by default (see Visual effects tab to enable).

Whatu · Sep 18, 2008

Yes, self-shadowing will give orbiter a new dimension... very good work Artlav, keep it coming! . Are you doing this just by yourself?

Weirdo Earthtorch · Sep 18, 2008

Hi Artlav, thanks for this - finally got it up and running.

Only messed about briefly so far, noticed the following

- No base textures KSC as above {sorry crossed with your explanation}

- External view of some bodies causes CTD (eg Hyperion; Nereid)

{checked the log:-

18/09/2008 16:42:44| PLNT| Loading Textures/Nereid.tex OK
18/09/2008 16:42:44| ORBGL| Error: Error in drwoplanet (i=1)
etc...}

- Time acceleration -possibly when starting from an initial Ground Observer View causes CTD (tried DG Brighton Beach; Views -Martial Atmosphere Colours). Changing view mode to something else then back can sometimes fix.

{couldn't see anything untoward in orbiter.log}

- High shadow mode (via F7 menu) causes FPS to drop significantly (V high mode (NA?)fixes, but shadows can be 'blocky' & don't display properly - sorry don't have any hosting to upload a pic.)

- FPS 2008 edition/scripts/space shuttle launch = 20 @1680x1050x32 windowed. Fullscreen (20FPS) always seems to get overridden into 1680x1050x24

Hope this info is useful - I'll try to test some more

Cheers W.E.

P4 3.4
2Gb RAM
Geforce6800 256Mb NForce 93.71
XP Home

Artlav · Sep 18, 2008

Whatu said:
Are you doing this just by yourself?

Sure, OGLAClient being just about 50000 lines it's not too complicated to warrant a team, and anyway most of the time it takes is thinking about how to get all the pieces and interactions together - not a parallelizeable task.

Weirdo Earthtorch said:
- No base textures KSC as above

Right, i'll get that one up to speed ASAP.

- External view of some bodies causes CTD (eg Hyperion; Nereid)

Confirmed, investigating. Seems to be something about meshed planetoids.

- Time acceleration -possibly when starting from an initial Ground Observer View causes CTD (tried DG Brighton Beach; Views -Martial Atmosphere Colours). Changing view mode to something else then back can sometimes fix.

Could not reproduce - Martian Atmosphere Colours scenario just spins, no crash. Neither is Brighton Beach.
A sane OGLA crash should leave a couple dozen lines in the log and specific error message.

btw, are you using OGLA 080918-C and beta 080918?

- High shadow mode (via F7 menu) causes FPS to drop significantly (V high mode (NA?)fixes, but shadows can be 'blocky' & don't display properly)

Quite right. The Low, Medium and High modes uses precise, yet very heavy algorithm, so they are good for simple scenes.
VHigh uses an approximation algorithm, so is much faster on complex scenes, yet looks worse (also because it is not quite perfected yet).
I hope to make an auto selector that will switch shadow modes based on scene complexity to fit best performance.

- FPS 2008 edition/scripts/space shuttle launch = 20 @1680x1050x32 windowed. Fullscreen (20FPS) always seems to get overridden into 1680x1050x24

Um. Didn't quite get the problem.
If you speak about FPS, please specify what settings you use, if you speak about a problem, please clarify.

Weirdo Earthtorch · Sep 18, 2008

Hi Artlav,

btw, are you using OGLA 080918-C and beta 080918?

Yes. (I couldn't actually get the previous versions to work, although that might be my own incompetence)

Quote:
- FPS 2008 edition/scripts/space shuttle launch = 20 @1680x1050x32 windowed. Fullscreen (20FPS) always seems to get overridden into 1680x1050x24
Um. Didn't quite get the problem.
If you speak about FPS, please specify what settings you use, if you speak about a problem, please clarify.

My apologies. This was just performance feedback, not a problem.

I will try and reproduce the other problems too.

Artlav · Sep 19, 2008

Update, 080919.
Downloadable here:
http://orbides.1gb.ru/orbf/oglaclient-080919_beta-080918.zip (1.0 Mb)
For beta 080918.

New:
*Many small fixes
*Lv9+ textures support restored
*Orulex: Evolution, as i called the now cleaned and reworked terrain system is turned on by default. Here goes the main test point.

Global Heightmaps for terrain support:
http://orbides.1gb.ru/orbf/earth-hmap-lv8.zip (84Mb)
http://orbides.1gb.ru/orbf/mars-hmap-lv8.zip (82Mb)
http://orbides.1gb.ru/orbf/moon-hmap-lv8.zip (81Mb)

As was just said, the planetary visuals on the ground and from space is the testing priority of this version, if you are going to test it, i want to know, whether it looks good/acceptable in motion and in static, from different altitudes and locations, as well as if it works fast enough.

The system is adaptive-multithreaded now, meaning that it adjusts itself to the number of CPU cores present on the machine. In any case, it is expected to run fluidly on dual- and more- core machines and acceptably on single-core machines.
Whatever there is, performance reports are welcome.

The baseline numbers are expected to vary, here it takes 4 seconds to reach Lv14 in Cruising above Florida scenario, and generally under 2 seconds to get up to Lv11 in orbit after half-round-the-Earth time accel.

What is reasonable to expect:
Orbital:

KSC Update:

Cruising above Florida:

Close-up on Shuttle:

Grawp · Sep 20, 2008

hmm. I can't find orbiter beta thread in this new forum. could somebody give me a link pls?

New Release OGLA OpenGL Client 080919

Aperiodic traveller

O-F Administrator

Addon Developer

Aperiodic traveller

Addon Developer

Aperiodic traveller

Addon Developer

Aperiodic traveller

Addon Developer

Aperiodic traveller

Addon Developer

Aperiodic traveller

Addon Developer

Attachments

Aperiodic traveller

Interplanetary Stowaway

Member

Aperiodic traveller

Member

Aperiodic traveller

New member

Similar threads