0.09 Alpha 7 Release

Developer's Forum, for discussion of bugs, code, and other developmental aspects of OpenGLide.

Postby Glidos » 2003-1-05 @ 10:40

CraigG wrote:Paul, any chance you could test out
Ignition
Croc
Outlaws

for me, just to make sure it ain't my PC.

I doubt it is, as it does this in WinME, Win2k, and WinXP.

Yes, I hope to very soon, but I don't think we should expect any changes in these games from the latest releases. There is a long-standing transparency problem, and Fabio wasn't specifically looking to cure that; he was looking at optimisations. He might fix it by accident of course, while simplifying program logic, but it looks like he hasn't. (All that's assuming you're seeing only a few transparency problems in Screamer 2).

Also the speed increases were not in the LFB bottleneck, so we shouldn't expect improvement in Screamer.

The remaining known transparency problems are next on my list.
User avatar
Glidos
l33t
 
Posts: 3107
Joined: 2002-8-08 @ 02:55

Postby Glidos » 2003-1-05 @ 10:43

Kaminari wrote:Fabio, still no luck with the Inventory screen :-/

Paul, I've tested Glidos 1.19 (from the original archive) with same result. Incidentally, I've noticed that Lara's shadow disappears with Glidos 1.19 + OpenGlide 0.09rc1, but is correctly displayed with Glidos 1.19b3 (Blood ver 3) + OpenGlide 0.09rc1. I took care of deleting the settings.dat file between the tests...

The mystery's thickening ;)

I've been working from CVS. I'll try grabbing the release candidate.
User avatar
Glidos
l33t
 
Posts: 3107
Joined: 2002-8-08 @ 02:55

Postby Glidos » 2003-1-05 @ 13:14

I've tried 009rc1 with all the Glidos games (except Carmageddon), and the Windows games that I usually test (RB3D, SpecOps, Sentinel Returns). There is only one graphical difference I can see, and that is in SpecOps.

Mostly the change in SpecOps is an improvement: the video sequences now play. Must be the LFB changes. But also the menu text has become messed up. I now get things like

Code: Select all
A AB BO OR RT T   G GA AM ME


With the letters jiggled about a bit in terms of their heights. Overall, I'd still say SpecOps has improved, so no results of mine suggest any reason not to make 009rc1 into 009.

Hind hasn't magically started working, previously existing transparency problems haven't gone away, and the Redguard sun is still odd, but that's as expected I think.
User avatar
Glidos
l33t
 
Posts: 3107
Joined: 2002-8-08 @ 02:55

Postby CraigG » 2003-1-05 @ 16:53

Yes, I hope to very soon, but I don't think we should expect any changes in these games from the latest releases. There is a long-standing transparency problem, and Fabio wasn't specifically looking to cure that; he was looking at optimisations. He might fix it by accident of course, while simplifying program logic, but it looks like he hasn't. (All that's assuming you're seeing only a few transparency problems in Screamer 2).

Also the speed increases were not in the LFB bottleneck, so we shouldn't expect improvement in Screamer.

The remaining known transparency problems are next on my list.


Yeah, I'm not too worried to be honest.
The speed increases are nice for owners of older PCs, I guess. But I'm already running everything way too fast anyway, so I don't see the increases soo much.
I think he's getter there, though. :)
Athlon 64 3000+ stock
MSI NForce 4 K8N Neo Platinum
2Gb RAM
nVidia Geforce 6800GT stock clocks
SBLive! Platinum + Audigy ZS2 Drivers
WinXP Pro SP2
User avatar
CraigG
Member
 
Posts: 207
Joined: 2002-9-01 @ 16:13

Postby Snover » 2003-1-06 @ 00:34

It sounds like a framerate limiter should be added before 009.
Snover
l33t++
 
Posts: 5204
Joined: 2002-6-30 @ 04:47

Postby Glidos » 2003-1-06 @ 00:57

Snover wrote:It sounds like a framerate limiter should be added before 009.

How many games do we have where too high a frame rate is a problem? I thought most games based their timing on other things.

Setting vertical sync in your OpenGL setup isn't a bad idea. As well as limiting frame rate, it gets rid of horizontal tearing.
User avatar
Glidos
l33t
 
Posts: 3107
Joined: 2002-8-08 @ 02:55

New RC2

Postby GoldFinger » 2003-1-07 @ 19:02

New 0.09 Rc2 on the SourceForge, please try it on... :)
GoldFinger
Member
 
Posts: 152
Joined: 2002-9-26 @ 17:00

Fixed the transparency problem

Postby Glidos » 2003-1-07 @ 21:48

I think I've fixed the transparency problems. At least
Screamer 2 and Extreme Assault are working correctly now.
Haven't tried Croc yet.

Shall I release it as 009rc3?

That's assuming rc2 doesn't already fix it. I've only just discovered that rc2 exists.
User avatar
Glidos
l33t
 
Posts: 3107
Joined: 2002-8-08 @ 02:55

Re: Fixed the transparency problem

Postby Kaminari » 2003-1-07 @ 22:36

Glidos wrote:Shall I release it as 009rc3?


Yes please ^^
User avatar
Kaminari
Oldbie
 
Posts: 681
Joined: 2002-12-06 @ 01:54

Great

Postby GoldFinger » 2003-1-07 @ 22:43

Great mews, what was the problem?
Merge my changes with yours and release it as 0.09rc3, if everything is ok we will release it as 0.09.
GoldFinger
Member
 
Posts: 152
Joined: 2002-9-26 @ 17:00

Postby Glidos » 2003-1-07 @ 23:56

Okay it should be there now. Chromakeying previously worked only for P8 textures. I've just extended it to 565 textures. We may need to extend it further, but I think this will be enough to catch a lot of the problems.
User avatar
Glidos
l33t
 
Posts: 3107
Joined: 2002-8-08 @ 02:55

Postby CraigG » 2003-1-08 @ 01:49

You loverly person, you :)

Looked what you fixed :D

Only one tiny, weeny little thing.
What happened to windowed mode ?
i.e. When in windowed mode, if you click outside of the game (say the desktop), the window minimises.

It's only a tiny thing, but I sometimes play around with gamma while the game is running. (Nvidia drivers too dark at times)

But Paul, if you were female, I'd give you a huge sloppy kiss right now ;)
Attachments
croc.jpg
croc.jpg (24.5 KiB) Viewed 2259 times
Athlon 64 3000+ stock
MSI NForce 4 K8N Neo Platinum
2Gb RAM
nVidia Geforce 6800GT stock clocks
SBLive! Platinum + Audigy ZS2 Drivers
WinXP Pro SP2
User avatar
CraigG
Member
 
Posts: 207
Joined: 2002-9-01 @ 16:13

Postby Kaminari » 2003-1-08 @ 01:52

Double ten thumbs up, guys. Now we're talking serious!

Every complaint of mine regarding TR has been resolved (except for that TRUB audio CD oddity ;) ). I hope to grab a copy of Redguard and give it a try soon. I really want to have a closer look at those Elder Scrolls antiquities before digging into Morrowind :D

Back to the Hive!

Edit: for what I can see of it, I might as well get my hands on Croc! I really enjoyed the early PSX demo, but for some reason I just forgot it somewhere on the long road towards TR5 :P
User avatar
Kaminari
Oldbie
 
Posts: 681
Joined: 2002-12-06 @ 01:54

Postby Glidos » 2003-1-08 @ 09:04

Ah, great. I thought that might work. Not sure what's changed with windowed mode. Nothing I've altered as far as I know.

Also, this change shouldn't have made a difference to TR1, but still if it seems to be working okay now, all well and good.
User avatar
Glidos
l33t
 
Posts: 3107
Joined: 2002-8-08 @ 02:55

Postby GoldFinger » 2003-1-08 @ 11:11

The problem with TR1 was a bug I did with the color combine modes, I fixed that and you fixed a lot of games... :)
OpenGLide is becoming the better wrapper of all... :P
GoldFinger
Member
 
Posts: 152
Joined: 2002-9-26 @ 17:00

no MMX, out!!!!

Postby GoldFinger » 2003-1-08 @ 19:08

Well,
MMX is the standard these days, all CPUs in use has them so I think we can put MMX as a requirement in OpenGLide, we check for its existence and them warns the user if the CPU has no support for it, but damn, we cannot run OpenGLide in anything less than an k6-2 in our days and the only CPUs that do not support MMX are the P200 and less, what do you all think? I could remove a lot of "if" in OpenGLide and make the code simpler.
GoldFinger
Member
 
Posts: 152
Joined: 2002-9-26 @ 17:00

Postby Kaminari » 2003-1-08 @ 20:42

I've got a P3/620 and a P2/420 (you hear me right!), which must be two of the lowest CPUs on this forum ever, so no MMX problem on my side :D
User avatar
Kaminari
Oldbie
 
Posts: 681
Joined: 2002-12-06 @ 01:54

Re: no MMX, out!!!!

Postby Glidos » 2003-1-08 @ 22:56

GoldFinger wrote:Well,
MMX is the standard these days, all CPUs in use has them so I think we can put MMX as a requirement in OpenGLide, we check for its existence and them warns the user if the CPU has no support for it, but damn, we cannot run OpenGLide in anything less than an k6-2 in our days and the only CPUs that do not support MMX are the P200 and less, what do you all think? I could remove a lot of "if" in OpenGLide and make the code simpler.


We could have a virtual base class with two derived classes one for the MMX implementation; one for the non-MMX. That would then be very clean, and I don't think the virtual call overhead would be anywhere near significant.
User avatar
Glidos
l33t
 
Posts: 3107
Joined: 2002-8-08 @ 02:55

Postby Glidos » 2003-1-08 @ 22:59

GoldFinger wrote:The problem with TR1 was a bug I did with the color combine modes, I fixed that and you fixed a lot of games... :)
OpenGLide is becoming the better wrapper of all... :P

Actually my change was very small. It is you that keeps making the big leaps. Great to see the development going so well.
User avatar
Glidos
l33t
 
Posts: 3107
Joined: 2002-8-08 @ 02:55

Postby twins73 » 2003-1-11 @ 10:31

Well, I am not from your development team or poll of testers so
ignore my comments if I am not supposed to bug you guys :D

I hate to be a bearer of bad news so hopefully all is fine and I
am wrong in some way (surely :)).

When I replace openglide 9a4 with 9rc3 I have the following message
that pops out

Application popup: Glidos.exe - Application Error : The instruction at "0x1000ad2f" referenced memory at "0xcfe6c1f4". The memory could not be "read".

This appeared to be for all version of openglide starting at 9a7.
9a6 runs fine (well with its transperancies issues) but starting from
9a7 my glidos blows.

So well, I really wanted to try 9rc3 so sorry to bring bad news.

My config is : P4 2.53Ghz, 512 Mb DDR2700, Asus P4PE-SATA, GF4-Ti4200, Win 2K-SP3, DX9 (<- oups :confused: )
My test : Playing Unfinished Business with Glidos 1.19.1

So would you guys be able to trace something common to 9a7,
9a8, 9rc1, 9rc2 and 9rc3 that makes them different to 9a6, 9a5, 9a4, eteceterat ?

Sorry for the disturbance. I just hope this help before you release
a final 9 and others start to (who knows) bring this up too.

P.S.: BTW Paul, thanks again for that wonderful Glidos of yours

Just thought those files might help ?

<OpenGLid.LOG>

--------------------------------------------------------
OpenGLide Log File
--------------------------------------------------------
***** OpenGLide 0.09rc3 *****
--------------------------------------------------------
Date: 01/11/03
Time: 10:55:37
--------------------------------------------------------
--------------------------------------------------------
Clock Frequency: 2539.10 Mhz
--------------------------------------------------------
--------------------------------------------------------
Configuration file is OpenGLid.INI
--------------------------------------------------------
Wrapper Class Priority of 32
Wrapper Priority of 2
--------------------------------------------------------
ColorBits = 16
DepthBits = 16
--------------------------------------------------------
** OpenGL Information **
--------------------------------------------------------
Vendor: NVIDIA Corporation
Renderer: GeForce4 Ti 4200/AGP/SSE2
Version: 1.4.0
Extensions: GL_ARB_depth_texture GL_ARB_imaging GL_ARB_multisample GL_ARB_multitexture GL_ARB_point_parameters GL_ARB_shadow GL_ARB_texture_border_clamp GL_ARB_texture_compression GL_ARB_texture_cube_map GL_ARB_texture_env_add GL_ARB_texture_env_combine GL_ARB_texture_env_dot3 GL_ARB_texture_mirrored_repeat GL_ARB_transpose_matrix GL_ARB_vertex_program GL_ARB_window_pos GL_S3_s3tc GL_EXT_abgr GL_EXT_bgra GL_EXT_blend_color GL_EXT_blend_minmax GL_EXT_blend_subtract GL_EXT_compiled_vertex_array GL_EXT_draw_range_elements GL_EXT_fog_coord GL_EXT_multi_draw_arrays GL_EXT_packed_pixels GL_EXT_paletted_texture GL_EXT_point_parameters GL_EXT_rescale_normal GL_EXT_secondary_color GL_EXT_separate_specular_color GL_EXT_shadow_funcs GL_EXT_shared_texture_palette GL_EXT_stencil_wrap GL_EXT_texture3D GL_EXT_texture_compression_s3tc GL_EXT_texture_cube_map GL_EXT_texture_edge_clamp GL_EXT_texture_env_add GL_EXT_texture_env_combine GL_EXT_texture_env_dot3 GL_EXT_texture_filter_anisotropic GL_EXT_texture_lod GL_EXT_texture_lod_bias GL_EXT_texture_object GL_EXT_vertex_array GL_EXT_vertex_weighting GL_HP_occlusion_test GL_IBM_texture_mirrored_repeat GL_KTX_buffer_region GL_NV_blend_square GL_NV_copy_depth_to_color GL_NV_depth_clamp GL_NV_fence GL_NV_fog_distance GL_NV_light_max_exponent GL_NV_multisample_filter_hint GL_NV_occlusion_query GL_NV_packed_depth_stencil GL_NV_pixel_data_range GL_NV_point_sprite GL_NV_register_combiners GL_NV_register_combiners2 GL_NV_texgen_reflection GL_NV_texture_compression_vtc GL_NV_texture_env_combine4 GL_NV_texture_rectangle GL_NV_texture_shader GL_NV_texture_shader2 GL_NV_texture_shader3 GL_NV_vertex_array_range GL_NV_vertex_array_range2 GL_NV_vertex_program GL_NV_vertex_program1_1 GL_SGIS_generate_mipmap GL_SGIS_multitexture GL_SGIS_texture_lod GL_SGIX_depth_texture GL_SGIX_shadow GL_WIN_swap_hint WGL_EXT_swap_control
--------------------------------------------------------
OpenGL Extensions:
--------------------------------------------------------
Extension GL_EXT_secondary_color is present and ENABLED
Extension GL_ARB_multitexture is present and ENABLED
Extension GL_EXT_fog_coord is present and ENABLED
Extension GL_EXT_texture_env_add is present and ENABLED
Extension GL_EXT_texture_env_combine is present and ENABLED
Extension GL_EXT_texture_lod_bias is present and ENABLED
Note: OpenGL GL_SGIS_generate_mipmap extension is supported but disabled by user
Extension GL_EXT_paletted_texture is present and ENABLED
Note: OpenGL GL_EXT_vertex_array extension is supported but disabled by user
Note: OpenGL GL_EXT_blend_func_separate extension is not supported, emulating behavior.
--------------------------------------------------------
MultiTexture Textures Units = 4
Using Palette Extension.
--------------------------------------------------------
Setting in Use:
--------------------------------------------------------
Init Full Screen = false
Fog = true
Precision Fix = true
Wrap 565 to 5551 = true
Texture Memory Size = 16 Mb
Frame Buffer Memory Size = 8 Mb
MMX is enabled
--------------------------------------------------------
--------------------------------------------------------
** Glide Calls **
--------------------------------------------------------

</OpenGLid.LOG>

<OpenGLid.INI>

Configuration File for OpenGLide

Info:
Priority goes from 0(HIGH) to 5(IDLE)
Texture Memory goes from 2 to 32
Frame Buffer Memory goes from 2 to 16
All other fields are boolean with 1(TRUE) and 0(FALSE)

Version=0.09rc3

[Options]
WrapperPriority=2
CreateWindow=0
InitFullScreen=0
EnableMipMaps=0
IgnorePaletteChange=0
Wrap565to5551=1
EnablePrecisionFix=1
EnableMultiTextureEXT=1
EnablePaletteEXT=1
EnableVertexArrayEXT=0
TextureMemorySize=16
FrameBufferMemorySize=8

</OpenGLid.INI>

<Glidos.ini>

...

Selected: Unfinished Business

Name: Unfinished Business
Executable: H:\Tombraid\Tombub.exe
ExeSizes: 867563, 868075
DosGraphics: Yes
VESASupport: Yes
ControlJudderFix: 100
Resolution: 2
FullScreen: No

</Glidos.ini>

Sorry for this lengthly post
twins73
Newbie
 
Posts: 8
Joined: 2002-12-20 @ 19:26

PreviousNext

Return to OpenGLide Development

Who is online

Users browsing this forum: No registered users and 0 guests