|
#4514944 - 04/06/20 12:28 PM
Re: EECH Source Code Discussion
[Re: messyhead]
|
Joined: Mar 2020
Posts: 15
D4rthCoffee
Junior Member
|
Junior Member
Joined: Mar 2020
Posts: 15
Bucks, UK
|
Would you mind moving your branch onto the main repo? I'd like to try merging it with my branch and start using VS from now on.
The merge is done.
As long as thealx is ok with it too, and happy for Ogre to have been removed (although it will still be available on other branches), and it can be merged into master once we know it's all working ok, with on issues. After 1.16.1 is released, I'll make a branch off master as the latest branch with Ogre and Watcom to keep as an archive.
Would you also be ok with adding a wiki page about using VS?
Will do, no problem. I can update the Watcom page to mention that it doesn't apply to vs2019 branch or its derivatives. Cheers
Last edited by D4rthCoffee; 04/06/20 12:53 PM.
|
|
|
#4514952 - 04/06/20 02:11 PM
Re: EECH Source Code Discussion
[Re: tulyopt]
|
Joined: Dec 2010
Posts: 1,878
messyhead
Member
|
Member
Joined: Dec 2010
Posts: 1,878
|
Are you able to play the game after compiling it with VS? I merged vs2019 into my branch, but when I try to enter a cockpit, the game hangs. This is the last entry in the debug log
DYNAMICS: ---------------------------------- Landed ---------------------------------------------- @ 61.80
DYNAMICS: added flight path data [138286.921875, 64694.234375] action ACTION_USER_NAVIGATING @ 61.80
DYNAMICS: model within keysite area, repairing and refueling @ 61.80
Failed to locate a terrain face (x = nan, z = nan, x_sector = -2147483648, z_sector = -2147483648, x_offset = nan, z_offset = nan @ 65.00
Failed to locate a terrain face (x = nan, z = nan, x_sector = -2147483648, z_sector = -2147483648, x_offset = nan, z_offset = nan @ 65.00
Failed to locate a terrain face (x = nan, z = nan, x_sector = -2147483648, z_sector = -2147483648, x_offset = nan, z_offset = nan @ 65.00
Failed to locate a terrain face (x = nan, z = nan, x_sector = -2147483648, z_sector = -2147483648, x_offset = nan, z_offset = nan @ 65.00
************************************ FATAL ***********************************
ASSERT x >= 0 (D:\eech_1_16_1_blackhawk_cockpit\modules\3d\terrain\terrdata.c:1697)
|
|
|
#4515036 - 04/06/20 10:24 PM
Re: EECH Source Code Discussion
[Re: messyhead]
|
Joined: Mar 2020
Posts: 15
D4rthCoffee
Junior Member
|
Junior Member
Joined: Mar 2020
Posts: 15
Bucks, UK
|
Are you able to play the game after compiling it with VS? I merged vs2019 into my branch, but when I try to enter a cockpit, the game hangs. This is the last entry in the debug log
DYNAMICS: ---------------------------------- Landed ---------------------------------------------- @ 61.80
DYNAMICS: added flight path data [138286.921875, 64694.234375] action ACTION_USER_NAVIGATING @ 61.80
DYNAMICS: model within keysite area, repairing and refueling @ 61.80
Failed to locate a terrain face (x = nan, z = nan, x_sector = -2147483648, z_sector = -2147483648, x_offset = nan, z_offset = nan @ 65.00
Failed to locate a terrain face (x = nan, z = nan, x_sector = -2147483648, z_sector = -2147483648, x_offset = nan, z_offset = nan @ 65.00
Failed to locate a terrain face (x = nan, z = nan, x_sector = -2147483648, z_sector = -2147483648, x_offset = nan, z_offset = nan @ 65.00
Failed to locate a terrain face (x = nan, z = nan, x_sector = -2147483648, z_sector = -2147483648, x_offset = nan, z_offset = nan @ 65.00
************************************ FATAL ***********************************
ASSERT x >= 0 (D:\eech_1_16_1_blackhawk_cockpit\modules\3d\terrain\terrdata.c:1697) I have not tested the gameplay yet, but considering that the compiler is 100% strict, compared to Watcom, it would seem quite viable that it won't work just like that. I found tons of inconsistencies, where variable values are truncated or the precision is lost, and probably watcom wouldn't care, but VS compiles strictly according to C++ rules. I'll have a look at the gameplay, and check why would it fail, It could be simply due to nedmalloc not being used anymore.
|
|
|
#4515464 - 04/09/20 12:56 PM
Re: EECH Source Code Discussion
[Re: messyhead]
|
Joined: Mar 2020
Posts: 15
D4rthCoffee
Junior Member
|
Junior Member
Joined: Mar 2020
Posts: 15
Bucks, UK
|
Actually, now I think about it, I'm sure your vs2019 branch worked OK. So it's maybe something introduced when I merged it into my branch Some of the issues I found to be caused by "attitude" values for the moving collision to be NaN. I'm going to trace the root cause of this.
|
|
|
#4517044 - 04/18/20 10:53 AM
Re: EECH Source Code Discussion
[Re: tulyopt]
|
Joined: Dec 2012
Posts: 1,381
thealx
Member
|
Member
Joined: Dec 2012
Posts: 1,381
|
D4rthCoffee, I was able to compile EXE after a lot of experiments with different DX and Windows SDK versions. some warnings appear while compiling, but no errors.
so I can launch application and use main menu UI, but after I am trying to start campaign/mission - game crashed without any message. have you tested gameplay so far?
I can try to fix warnings but not sure it will help in my situation, as it can be wrong libs or VS configs etc.
Last edited by messyhead; 04/18/20 12:18 PM.
|
|
|
#4517053 - 04/18/20 12:20 PM
Re: EECH Source Code Discussion
[Re: thealx]
|
Joined: Dec 2010
Posts: 1,878
messyhead
Member
|
Member
Joined: Dec 2010
Posts: 1,878
|
D4rthCoffee, I was able to compile EXE after a lot of experiments with different DX and Windows SDK versions. some warnings appear while compiling, but no errors.
so I can launch application and use main menu UI, but after I am trying to start campaign/mission - game crashed without any message. have you tested gameplay so far?
I can try to fix warnings but not sure it will help in my situation, as it can be wrong libs or VS configs etc. That's what was happening to me as well. Although I was able to compile without issues straight from the vs2019 branch. I didn't get an error launching the game, but the error I posted above was in the debug logs.
|
|
|
#4517068 - 04/18/20 03:07 PM
Re: EECH Source Code Discussion
[Re: messyhead]
|
Joined: Dec 2012
Posts: 1,381
thealx
Member
|
Member
Joined: Dec 2012
Posts: 1,381
|
That's what was happening to me as well. Although I was able to compile without issues straight from the vs2019 branch. I didn't get an error launching the game, but the error I posted above was in the debug logs. looks like you get into the game, so you made better progress than me. I could give it another try from scratch. UPDATE: my bad, reason was in media files, somehow game install became broken after my experiments. After clean reinstall, it seems to work fine - can start the game and get inside helicopter. Will solve warnings then and make some tests.
Last edited by thealx; 04/19/20 12:41 PM.
|
|
|
#4518728 - 04/29/20 07:46 PM
Re: EECH Source Code Discussion
[Re: thealx]
|
Joined: Dec 2010
Posts: 1,878
messyhead
Member
|
Member
Joined: Dec 2010
Posts: 1,878
|
I've pushed some fixes, game playable now but some critical issue with textures initialization exists that I wasn't able to solve so far - like wrong textures for dynamic water feature, or cash on game close attempt in the middle of textures unloading process. That's good you've made progress with it.
|
|
|
#4518733 - 04/29/20 08:02 PM
Re: EECH Source Code Discussion
[Re: tulyopt]
|
**DONOTDELETE**
Unregistered
|
BANITA
Unregistered
|
Noob question
:if you solve all the problems , there are some graphics improvements by default? or do you need have to add everything and this is a long way?
|
|
|
#4518916 - 04/30/20 09:54 PM
Re: EECH Source Code Discussion
[Re: D4rthCoffee]
|
Joined: Dec 2012
Posts: 1,381
thealx
Member
|
Member
Joined: Dec 2012
Posts: 1,381
|
Texture issues fixed, game seems to be playable. Left commits as pull request as not sure I didn't break anything.
I am going to work in this branch for some time so maybe other issues will pop up while testing my own things.
D4rthCoffee, if you are still interested in code improvements, I have several questions for you:
1. how hard is implement multi-cpu calculations in the game. no synchronization needed - it can be used for precise weather calculations, radio waves tracing, AI flight model and other things that does not require time precision
2. in music testing process I have reached x86 RAM size limit (about 1.5GB), is it possible to compile x64 executable easily? I know some hacks exist for x86 to increase memory limits, but they affect stability in a bad way
3. I see you've removed OGRE code partially, does it mean we will remove all of it at the end? there is no way back - if game will not get DX upgrade, we will lose alternative as well.
|
|
|
#4518924 - 04/30/20 10:44 PM
Re: EECH Source Code Discussion
[Re: thealx]
|
Joined: Dec 2010
Posts: 1,878
messyhead
Member
|
Member
Joined: Dec 2010
Posts: 1,878
|
Texture issues fixed, game seems to be playable. Left commits as pull request as not sure I didn't break anything.
I am going to work in this branch for some time so maybe other issues will pop up while testing my own things.
D4rthCoffee, if you are still interested in code improvements, I have several questions for you:
1. how hard is implement multi-cpu calculations in the game. no synchronization needed - it can be used for precise weather calculations, radio waves tracing, AI flight model and other things that does not require time precision
2. in music testing process I have reached x86 RAM size limit (about 1.5GB), is it possible to compile x64 executable easily? I know some hacks exist for x86 to increase memory limits, but they affect stability in a bad way
3. I see you've removed OGRE code partially, does it mean we will remove all of it at the end? there is no way back - if game will not get DX upgrade, we will lose alternative as well. Hey that's good you got it playable. I might merge it with my own branch and use VS from now on and see if there's any issues. Do you think it would he worth releasing a test exe to more players to see if they find any issues too? As for your question about removing Ogre, was it ever really an alternative? It seemed only Firebird was developing it, and I don't think it ever got extensively play tested.
|
|
|
#4518964 - 05/01/20 09:22 AM
Re: EECH Source Code Discussion
[Re: messyhead]
|
Joined: Dec 2012
Posts: 1,381
thealx
Member
|
Member
Joined: Dec 2012
Posts: 1,381
|
Do you think it would he worth releasing a test exe to more players to see if they find any issues too? it may be helpful, but I think it's too early - I believe a lot of bugs will appear even after quick gameplay tests As for your question about removing Ogre, was it ever really an alternative? It seemed only Firebird was developing it, and I don't think it ever got extensively play tested. it became working solution once rendering code was adjusted to game engine. it's not playable but can be compiled and improved by someone with average programming knowledge. another question that it will take so much time to complete, especially for single person.
|
|
|
#4518988 - 05/01/20 12:16 PM
Re: EECH Source Code Discussion
[Re: tulyopt]
|
Joined: Dec 2010
Posts: 1,878
messyhead
Member
|
Member
Joined: Dec 2010
Posts: 1,878
|
I merged vs2019 into my blackhawk branch, and applied your last commits as patches. But I get compile errors about min/max used in some places, for example co_grstab.c
if (eo_azimuth > 0.0)
{
eo_azimuth = min (eo_azimuth, eo_max_azimuth - 0.01);
}
else
{
eo_azimuth = max (eo_azimuth, eo_min_azimuth + 0.01);
}
error C2672: 'min': no matching overloaded function found error C2672: 'max': no matching overloaded function found Did you get these? Is there some compiler settings I need to change?
|
|
|
#4518992 - 05/01/20 12:27 PM
Re: EECH Source Code Discussion
[Re: thealx]
|
Joined: Dec 2010
Posts: 1,878
messyhead
Member
|
Member
Joined: Dec 2010
Posts: 1,878
|
my very last master commits that I haven't applied to VS2
try to add "f" to the end of numbers, like
eo_azimuth = min(eo_azimuth, eo_max_azimuth - 0.01f); Ok, I'll try that. I also thought about using fmin/fmax.
|
|
|
|
