filipetolhuizen wrote on 2025-04-27, 00:57:

sharangad wrote on 2025-04-26, 23:17:

I just tested this on Win11/24H2 with the April updates and it still runs for me. […]
Show full quote

filipetolhuizen wrote on 2025-04-26, 16:05:

SODA Off Road Racing is crashing for me with the latest build. Just returns to desktop after the startup screen.

I just tested this on Win11/24H2 with the April updates and it still runs for me.

[EDIT] Win 10 (April 2025 updates) also appears to work[/EDIT]

Could you click the "zip logfiles" button in RLauncher and attach the highlighted zip file to a post? If you could also let me know your system specs via PM that would be useful (run DXDiag and save the results form within the app and PM me the txt file).

Also a screenshot of your Rendition settings panel in RLauncher.

Latest working version for me was from early February (both 03 and 05). The one from the 24th was already broken as well as every build after that. I'm attaching the log from the april build as well as dxdiag.txt.

If you set your resolution to 1920x1080x60 instead of 144 Hz does it run?

[EDIT3] Don't try this. RLauncher should be ok. It uses a slightly older version of the MD5 generator. It isn't affected [/EDIT3]
When you get a chance, as a quick test with the broken version on Windows 7 could you:
(Assuming you're on an Alpha release)
1) Run Rlauncher
2) Go to the SDK Examples tab
3) Click "Reset SDK Location" (if it's not greyed out).
4) Click The "Download" button which appears at the centre of the SDK Examples tab.

Does this work or does RLauncher crash?

I suspect the problem is in the Cryptographic libraries which is used to calculate the MD5 checksums of some textures. There was a resource leak and some CPU waste from re-creating the part of it which re-calculated the hash. I set it to reset and reuse the existing hash. I *think* that's what's causing the crash. The Reuse flag might not be working on Windows 7. RLauncher also uses it when downloading demos to verify that the file it's downloading isn't corrupted or malware. SODA Offroad doesn't actually use the MD5s.

This might be an incompatibility with Windows 7. I'll fix it if you can confirm that RLauncher crashes. If you can wait a while longer , I'll get you a build with the older MD5 code to see if it runs.

[EDIT] That's not it. The MD5 generator isn't used in SODA.
[EDIT2] Yes it's used. RLauncher doesn't use it[/EDIT2].

filipetolhuizen wrote on 2025-04-26, 16:05:

SODA Off Road Racing is crashing for me with the latest build. Just returns to desktop after the startup screen.

Could you try this on your Win7 machine (Untested Alpha):
https://nirvtek.com/downloads/RReady.Alpha.20250427.001.7z
MD5: 5a0d6a99fa88c181d5fb7a4a9e525ed5

- Possible fix for SODA Offroad crashing on Windows 7. This *might* be because of the crypto libraries resetting and reusing the same hash generator for texture MD5 checksums.
The "reusable" setting might not be working on Windows 7. So, the generator is destroyed and recreated every single time it's used.

Also if you could run the RLauncher test on Windows 7, it would be useful. This build updates the MD5 code in RLauncher.

The first CGL function disassembles to this, or at least the first part does:

Does the outs instruction set any flags. According to the docs it doesn't. I can't be sure this isn't garbage code, but it does appear to disassemble without any errors.
[EDIT] This is garbage.
[code]0: 48 dec eax
1: 6f outs dx,DWORD PTR ds:[esi]
2: 74 4c je 0x50
4: 69 6e 6b 20 43 47 4c imul ebp,DWORD PTR [esi+0x6b],0x4c474320
b: 20 52 65 and BYTE PTR [edx+0x65],dl
e: 76 20 jbe 0x30
10: 32 0d 0a 43 6f 70 xor cl,BYTE PTR ds:0x706f430a
16: 79 72 jns 0x8a
18: 69 67 68 74 20 28 43 imul esp,DWORD PTR [edi+0x68],0x43282074
1f: 29 20 sub DWORD PTR [eax],esp
21: 31 39 xor DWORD PTR [ecx],edi
23: 39 35 2d 31 39 39 cmp DWORD PTR ds:0x3939312d,esi
29: 36 20 43 72 and BYTE PTR ss:[ebx+0x72],al
2d: 65 61 gs popa
2f: 74 69 je 0x9a
31: 76 65 jbe 0x98
33: 20 54 65 63 and BYTE PTR [ebp+eiz*2+0x63],dl
37: 68 6e 6f 6c 6f push 0x6f6c6f6e
3c: 67 79 20 addr16 jns 0x5f
3f: 4c dec esp
40: 74 64 je 0xa6
42: 2e 20 41 6c and BYTE PTR cs:[ecx+0x6c],al
46: 6c ins BYTE PTR es:[edi],dx
47: 20 72 69 and BYTE PTR [edx+0x69],dh
4a: 67 68 74 73 20 72 addr16 push 0x72207374
50: 65 73 65 gs jae 0xb8
53: 72 76 jb 0xcb
55: 65 64 2e 0d 0a 43 47 gs fs cs or eax,0x4c47430a
5c: 4c
5d: 31 37 xor DWORD PTR [edi],esi
5f: 30 49 52 xor BYTE PTR [ecx+0x52],cl
62: 54 push esp
63: 2e 56 cs push esi
65: 52 push edx
66: 49 dec ecx
67: 0d 0a 1a 48 4f or eax,0x4f481a0a
6c: 54 push esp
6d: 32 00 xor al,BYTE PTR [eax]
6f: aa stos BYTE PTR es:[edi],al
70: 00 54 10 c5 add BYTE PTR [eax+edx*1-0x3b],dl
74: 83 01 00 add DWORD PTR [ecx],0x0
77: 00 00 add BYTE PTR [eax],al
79: 00 00 add BYTE PTR [eax],al
7b: 15 00 00 00 00 adc eax,0x0
80: 10 00 adc BYTE PTR [eax],al
82: 00 00 add BYTE PTR [eax],al
84: 00 00 add BYTE PTR [eax],al
86: 00 4c 00 00 add BYTE PTR [eax+eax*1+0x0],cl
8a: 00 02 add BYTE PTR [edx],al
8c: 00 00 add BYTE PTR [eax],al
8e: 00 7c 00 00 add BYTE PTR [eax+eax*1+0x0],bh
92: 00 6f 04 add BYTE PTR [edi+0x4],ch
95: 00 00 add BYTE PTR [eax],al
97: 50 push eax
98: 05 00 00 a8 05 add eax,0x5a80000
9d: 00 00 add BYTE PTR [eax],al
9f: f8 clc
a0: 40 inc eax
a1: 00 00 add BYTE PTR [eax],al
a3: 24 01 and al,0x1
a5: 00 00 add BYTE PTR [eax],al
a7: 51 push ecx
a8: 3f aas
a9: 00 00 add BYTE PTR [eax],al
ab: 51 push ecx
ac: 3f aas
ad: 00 00 add BYTE PTR [eax],al
af: 00 00 add BYTE PTR [eax],al
b1: 00 00 add BYTE PTR [eax],al
b3: 00 00 add BYTE PTR [eax],al
b5: 00 00 add BYTE PTR [eax],al
b7: 00 00 add BYTE PTR [eax],al
b9: 00 00 add BYTE PTR [eax],al
bb: 00 00 add BYTE PTR [eax],al
bd: 00 00 add BYTE PTR [eax],al
bf: 00 00 add BYTE PTR [eax],al
c1: 00 00 add BYTE PTR [eax],al
c3: 00 00 add BYTE PTR [eax],al
c5: 00 00 add BYTE PTR [eax],al
c7: a0 5f 01 00 00 mov al,ds:0x15f
cc: 00 01 add BYTE PTR [ecx],al
ce: 00 03 add BYTE PTR [ebx],al
d0: 20 00 and BYTE PTR [eax],al
d2: 00 01 add BYTE PTR [ecx],al
d4: 00 00 add BYTE PTR [eax],al
d6: 00 0b add BYTE PTR [ebx],cl
d8: 00 00 add BYTE PTR [eax],al
da: 00 00 add BYTE PTR [eax],al
dc: 00 00 add BYTE PTR [eax],al
de: 00 63 95 add BYTE PTR [ebx-0x6b],ah
e1: 00 00 add BYTE PTR [eax],al
e3: 00 00 add BYTE PTR [eax],al
e5: 03 00 add eax,DWORD PTR [eax]
e7: 05 20 00 00 0c add eax,0xc000020
ec: 00 00 add BYTE PTR [eax],al
ee: 00 0a add BYTE PTR [edx],cl
f0: 00 00 add BYTE PTR [eax],al
f2: 00 00 add BYTE PTR [eax],al
f4: 00 00 add BYTE PTR [eax],al
f6: 00 00 add BYTE PTR [eax],al
f8: 00 00 add BYTE PTR [eax],al
fa: 00 00 add BYTE PTR [eax],al
fc: 10 00 adc BYTE PTR [eax],al
fe: 00 00 add BYTE PTR [eax],al
100: 10 00 adc BYTE PTR [eax],al
102: 00 00 add BYTE PTR [eax],al
104: 10 00 adc BYTE PTR [eax],al
106: 00 00 add BYTE PTR [eax],al
108: 20 00 and BYTE PTR [eax],al
10a: 00 00 add BYTE PTR [eax],al
10c: 10 00 adc BYTE PTR [eax],al
10e: 00 00 add BYTE PTR [eax],al
110: 30 00 xor BYTE PTR [eax],al
112: 00 00 add BYTE PTR [eax],al
114: 10 00 adc BYTE PTR [eax],al
116: 00 00 add BYTE PTR [eax],al
118: 40 inc eax
119: 00 00 add BYTE PTR [eax],al
11b: 00 10 add BYTE PTR [eax],dl
11d: 00 00 add BYTE PTR [eax],al
11f: 00 50 00 add BYTE PTR [eax+0x0],dl
122: 00 00 add BYTE PTR [eax],al
124: 10 00 adc BYTE PTR [eax],al
126: 00 00 add BYTE PTR [eax],al
128: 60 pusha
129: 00 00 add BYTE PTR [eax],al
12b: 00 10 add BYTE PTR [eax],dl
12d: 00 00 add BYTE PTR [eax],al
12f: 00 70 00 add BYTE PTR [eax+0x0],dh
132: 00 00 add BYTE PTR [eax],al
134: 10 00 adc BYTE PTR [eax],al
136: 00 00 add BYTE PTR [eax],al
138: 80 00 00 add BYTE PTR [eax],0x0
13b: 00 10 add BYTE PTR [eax],dl
13d: 00 00 add BYTE PTR [eax],al
13f: 00 90 00 00 00 10 add BYTE PTR [eax+0x10000000],dl
145: 00 00 add BYTE PTR [eax],al
147: 00 a0 00 00 6a 0d add BYTE PTR [eax+0xd6a0000],ah
14d: 00 00 add BYTE PTR [eax],al
14f: 6a ad push 0xffffffad
151: 00 00 add BYTE PTR [eax],al
153: 00 10 add BYTE PTR [eax],dl
155: 00 00 add BYTE PTR [eax],al
157: 6a bd push 0xffffffbd
159: 00 00 add BYTE PTR [eax],al
15b: 00 10 add BYTE PTR [eax],dl
15d: 00 00 add BYTE PTR [eax],al
15f: 6a cd push 0xffffffcd
161: 00 00 add BYTE PTR [eax],al
163: 00 10 add BYTE PTR [eax],dl
165: 00 00 add BYTE PTR [eax],al
167: 6a dd push 0xffffffdd
169: 00 00 add BYTE PTR [eax],al
16b: 00 10 add BYTE PTR [eax],dl
16d: 00 00 add BYTE PTR [eax],al
16f: 6a ed push 0xffffffed
171: 00 00 add BYTE PTR [eax],al
173: 00 10 add BYTE PTR [eax],dl
175: 00 00 add BYTE PTR [eax],al
177: 6a fd push 0xfffffffd
179: 00 00 add BYTE PTR [eax],al
17b: 00 10 add BYTE PTR [eax],dl
17d: 00 00 add BYTE PTR [eax],al
17f: 6a 0d push 0xd
181: 01 00 add DWORD PTR [eax],eax
183: 00 10 add BYTE PTR [eax],dl
185: 00 00 add BYTE PTR [eax],al
187: 6a 1d push 0x1d
189: 01 00 add DWORD PTR [eax],eax
18b: 00 10 add BYTE PTR [eax],dl
18d: 00 00 add BYTE PTR [eax],al
18f: 6a 2d push 0x2d
191: 01 00 add DWORD PTR [eax],eax
193: 00 10 add BYTE PTR [eax],dl
195: 00 00 add BYTE PTR [eax],al
197: 6a 3d push 0x3d
199: 01 00 add DWORD PTR [eax],eax
19b: 63 05 00 00 06 63 arpl WORD PTR ds:0x63060000,ax
1a1: 67 6c ins BYTE PTR es:[di],dx
1a3: 74 69 je 0x20e
1a5: 72 00 jb 0x1a7
1a7: 00 12 add BYTE PTR [edx],dl
1a9: 63 67 6c arpl WORD PTR [edi+0x6c],sp
1ac: 56 push esi
1ad: 65 72 73 gs jb 0x223
1b0: 69 6f 6e 43 6f 6e 74 imul ebp,DWORD PTR [edi+0x6e],0x746e6f43
1b7: 72 6f jb 0x228
1b9: 6c ins BYTE PTR es:[edi],dx
1ba: 5f pop edi
1bb: 01 00 add DWORD PTR [eax],eax
1bd: 0e push cs
1be: 63 67 6c arpl WORD PTR [edi+0x6c],sp
1c1: 53 push ebx
1c2: 65 6e outs dx,BYTE PTR gs:[esi]
1c4: 64 53 fs push ebx
1c6: 74 72 je 0x23a
1c8: 65 61 gs popa
1ca: 6d ins DWORD PTR es:[edi],dx
1cb: 5f pop edi
1cc: 02 00 add al,BYTE PTR [eax]
1ce: 0c 63 or al,0x63
1d0: 67 6c ins BYTE PTR es:[di],dx
1d2: 47 inc edi
1d3: 65 74 50 gs je 0x226
1d6: 69 78 65 6c 5f 03 00 imul edi,DWORD PTR [eax+0x65],0x35f6c
1dd: 0c 63 or al,0x63
1df: 67 6c ins BYTE PTR es:[di],dx
1e1: 50 push eax
1e2: 75 74 jne 0x258
1e4: 50 push eax
1e5: 69 .byte 0x69
1e6: 78 65 js 0x24d
1e8: 6c ins BYTE PTR es:[edi],dx
1e9: 5f pop edi
[/b][/code]

sharangad wrote on 2025-04-27, 04:45:

Could you try this on your Win7 machine (Untested Alpha): https://nirvtek.com/downloads/RReady.Alpha.20250427.001.7z MD5: 5a0d6 […]
Show full quote

filipetolhuizen wrote on 2025-04-26, 16:05:

SODA Off Road Racing is crashing for me with the latest build. Just returns to desktop after the startup screen.

Could you try this on your Win7 machine (Untested Alpha):
https://nirvtek.com/downloads/RReady.Alpha.20250427.001.7z
MD5: 5a0d6a99fa88c181d5fb7a4a9e525ed5

- Possible fix for SODA Offroad crashing on Windows 7. This *might* be because of the crypto libraries resetting and reusing the same hash generator for texture MD5 checksums.
The "reusable" setting might not be working on Windows 7. So, the generator is destroyed and recreated every single time it's used.

Also if you could run the RLauncher test on Windows 7, it would be useful. This build updates the MD5 code in RLauncher.

This build seems to be working fine with SODA, both standalone .dlls and RLauncher.

filipetolhuizen wrote on 2025-04-27, 22:05:

sharangad wrote on 2025-04-27, 04:45:

Could you try this on your Win7 machine (Untested Alpha): https://nirvtek.com/downloads/RReady.Alpha.20250427.001.7z MD5: 5a0d6 […]
Show full quote

filipetolhuizen wrote on 2025-04-26, 16:05:

SODA Off Road Racing is crashing for me with the latest build. Just returns to desktop after the startup screen.

Could you try this on your Win7 machine (Untested Alpha):
https://nirvtek.com/downloads/RReady.Alpha.20250427.001.7z
MD5: 5a0d6a99fa88c181d5fb7a4a9e525ed5

- Possible fix for SODA Offroad crashing on Windows 7. This *might* be because of the crypto libraries resetting and reusing the same hash generator for texture MD5 checksums.
The "reusable" setting might not be working on Windows 7. So, the generator is destroyed and recreated every single time it's used.

Also if you could run the RLauncher test on Windows 7, it would be useful. This build updates the MD5 code in RLauncher.

This build seems to be working fine with SODA, both standalone .dlls and RLauncher.

That's good to hear! I'll incorporate it into the next store release but it won't be soon. Even though this is an untested release, it should be almost identical to the last stable Alpha release.

Thanks for the update @felipetolhuizen!

sharangad wrote on 2025-04-28, 00:46:

filipetolhuizen wrote on 2025-04-27, 22:05:

sharangad wrote on 2025-04-27, 04:45:

Could you try this on your Win7 machine (Untested Alpha): https://nirvtek.com/downloads/RReady.Alpha.20250427.001.7z MD5: 5a0d6 […]
Show full quote

Could you try this on your Win7 machine (Untested Alpha):
https://nirvtek.com/downloads/RReady.Alpha.20250427.001.7z
MD5: 5a0d6a99fa88c181d5fb7a4a9e525ed5

- Possible fix for SODA Offroad crashing on Windows 7. This *might* be because of the crypto libraries resetting and reusing the same hash generator for texture MD5 checksums.
The "reusable" setting might not be working on Windows 7. So, the generator is destroyed and recreated every single time it's used.

Also if you could run the RLauncher test on Windows 7, it would be useful. This build updates the MD5 code in RLauncher.

This build seems to be working fine with SODA, both standalone .dlls and RLauncher.

That's good to hear! I'll incorporate it into the next store release but it won't be soon. Even though this is an untested release, it should be almost identical to the last stable Alpha release.

Thanks for the update @felipetolhuizen!

Anytime. I also tested it with Psygnosis Formula 1 and it's also working, although too fast in the menus and sluggish in races.

filipetolhuizen wrote on 2025-04-28, 02:38:

sharangad wrote on 2025-04-28, 00:46:

filipetolhuizen wrote on 2025-04-27, 22:05:

This build seems to be working fine with SODA, both standalone .dlls and RLauncher.

That's good to hear! I'll incorporate it into the next store release but it won't be soon. Even though this is an untested release, it should be almost identical to the last stable Alpha release.

Thanks for the update @felipetolhuizen!

Anytime. I also tested it with Psygnosis Formula 1 and it's also working, although too fast in the menus and sluggish in races.

How sluggish is it? The framerate should vary between 30-60 fps.

Is that what you mean or is it worse? The other rredline games should be full speed, except for "die by the sword."

sharangad wrote on 2025-04-28, 03:08:

filipetolhuizen wrote on 2025-04-28, 02:38:

sharangad wrote on 2025-04-28, 00:46:

That's good to hear! I'll incorporate it into the next store release but it won't be soon. Even though this is an untested release, it should be almost identical to the last stable Alpha release.

Thanks for the update @felipetolhuizen!

Anytime. I also tested it with Psygnosis Formula 1 and it's also working, although too fast in the menus and sluggish in races.

How sluggish is it? The framerate should vary between 30-60 fps.

Is that what you mean or is it worse? The other rredline games should be full speed, except for "die by the sword."

It stays at 144 in the menus but drops to 20-6 during the races.

filipetolhuizen wrote on 2025-04-28, 03:43:

sharangad wrote on 2025-04-28, 03:08:

filipetolhuizen wrote on 2025-04-28, 02:38:

Anytime. I also tested it with Psygnosis Formula 1 and it's also working, although too fast in the menus and sluggish in races.

How sluggish is it? The framerate should vary between 30-60 fps.

Is that what you mean or is it worse? The other rredline games should be full speed, except for "die by the sword."

It stays at 144 in the menus but drops to 20-6 during the races.

20.6?!?!?! The framerate in the game does vary, but I didn't think it'd go that slow. Is this on the same machine as SODA? I think Intel quad core with relatively modern nvidia graphics.

I'll check on my i7 2700. The translation from RRedline 1.0 to 2.0 does use the x86 instructions pushd and popd which are slow on modern Intel CPUs. It was very quick on older Pentiums /486s/386s and on older AMD FX CPUs (1-cycle, I think, ridiculously fast) , bit slower on Ryzen but still a lot faster than Intel. I think Intel doesn't optimise pushd and popd because they're not used very much, even in the 90s. Unfortunately RReady's RRedline 1.0 (v2k.RReady for running F1 on V2000 boards) uses it because it's the fastest way of saving all the registers. Raymond Chen from Microsoft mentioned those were ridiculously rarely used instructions on his blog.

I get 30-60 fps in F1 on my i7 2600. There is a moment where it's at 20 fps at the very start of the race. I tested a demo and an actual race, but the performance can vary depending on the track and the weather conditions.

You could try:
1) Capping the refresh rate to 60 Hz (select the target resolution with a 60 Hz refresh instead of selecting max)
2) Check that your machine is set to high performance mode in control panel/power and do the same for the GPU in nvidia control panel.

sharangad wrote on 2025-04-28, 04:22:

I get 30-60 fps in F1 on my i7 2600. There is a moment where it's at 20 fps at the very start of the race. I tested a demo and a […]
Show full quote

I get 30-60 fps in F1 on my i7 2600. There is a moment where it's at 20 fps at the very start of the race. I tested a demo and an actual race, but the performance can vary depending on the track and the weather conditions.

You could try:
1) Capping the refresh rate to 60 Hz (select the target resolution with a 60 Hz refresh instead of selecting max)
2) Check that your machine is set to high performance mode in control panel/power and do the same for the GPU in nvidia control panel.

My profiling options in this game are non-existent. Both VS 2022 and Intel vtune don't work with f1red.exe, the game just crashes. I think it's because of the register based calling convention.

Does this say anything for the Team Apache lockup situation:
https://github.com/narzoul/DDrawCompat/issues … ment-2749332168

BEEN_Nath_58 wrote on 2025-04-28, 09:10:

Does this say anything for the Team Apache lockup situation:
https://github.com/narzoul/DDrawCompat/issues … ment-2749332168

That sounds like the problem I'm having and my test machines are nvidia based. I can't test Intel based machines because front buffer based rendering doesn't work reliably and Team Apache's menus and videos use front buffer based rendering. I don't have any AMD GPUs either.

Maybe I should raise a ticket with nvidia. I don't know whether they'll fix it though. I don't think they target retro games and the userbase is kind of small.

Like Narzoul said, maybe I can lower priorities for all game threads on a separate wrapper thread continuously for TA.

sharangad wrote on 2025-04-28, 04:04:

filipetolhuizen wrote on 2025-04-28, 03:43:

sharangad wrote on 2025-04-28, 03:08:

How sluggish is it? The framerate should vary between 30-60 fps.

Is that what you mean or is it worse? The other rredline games should be full speed, except for "die by the sword."

It stays at 144 in the menus but drops to 20-6 during the races.

20.6?!?!?! The framerate in the game does vary, but I didn't think it'd go that slow. Is this on the same machine as SODA? I think Intel quad core with relatively modern nvidia graphics.

I'll check on my i7 2700. The translation from RRedline 1.0 to 2.0 does use the x86 instructions pushd and popd which are slow on modern Intel CPUs. It was very quick on older Pentiums /486s/386s and on older AMD FX CPUs (1-cycle, I think, ridiculously fast) , bit slower on Ryzen but still a lot faster than Intel. I think Intel doesn't optimise pushd and popd because they're not used very much, even in the 90s. Unfortunately RReady's RRedline 1.0 (v2k.RReady for running F1 on V2000 boards) uses it because it's the fastest way of saving all the registers. Raymond Chen from Microsoft mentioned those were ridiculously rarely used instructions on his blog.

Yes, same machine I run SODA, but SODA is running smoothly. CPU is I7 6700k and GPU, GTX 1660.

BEEN_Nath_58 wrote on 2025-04-28, 09:10:

Does this say anything for the Team Apache lockup situation:
https://github.com/narzoul/DDrawCompat/issues … ment-2749332168

That allows the videos to play without lock up. The first video plays back fine, the later ones have stuttery audio. The credits video freezes. The game works fine.

Changing priorities on thread #0 fixes the audio with a lengthy freeze between them.

I think the game thread and the asynchronous renderer thread need to have their priorities adjusted for this game. The others I think can be left alone.

BEEN_Nath_58 wrote on 2025-04-28, 09:10:

Does this say anything for the Team Apache lockup situation:
https://github.com/narzoul/DDrawCompat/issues … ment-2749332168

The fix in that thread worked, but not for all game threads, just the first 18 and only at the first switch to single threaded mode. I didn't realise system libraries like directsound used up so many threads.

Thanks @Been_Nath_58, this is a massive save.

*****************************************************
RReady Alpha 29 April 2025 (Untested Alpha Release)
*****************************************************
Team Apache Win11 24H2 fix.
https://nirvtek.com/downloads/RReady.Alpha.Te … 20250429.001.7z
MD5: 8e63dcdc883eaa769f3b83038ebfb88c

This should fix the campaign and credits videos stuttering as well.

The menu glitching is still there.

sharangad wrote on 2025-04-29, 12:36:

The fix in that thread worked, but not for all game threads, just the first 18 and only at the first switch to single threaded m […]
Show full quote

BEEN_Nath_58 wrote on 2025-04-28, 09:10:

Does this say anything for the Team Apache lockup situation:
https://github.com/narzoul/DDrawCompat/issues … ment-2749332168

The fix in that thread worked, but not for all game threads, just the first 18 and only at the first switch to single threaded mode. I didn't realise system libraries like directsound used up so many threads.

Thanks @Been_Nath_58, this is a massive save.

*****************************************************
RReady Alpha 29 April 2025 (Untested Alpha Release)
*****************************************************
Team Apache Win11 24H2 fix.
https://nirvtek.com/downloads/RReady.Alpha.Te … 20250429.001.7z
MD5: 8e63dcdc883eaa769f3b83038ebfb88c

This should fix the campaign and credits videos stuttering as well.

The menu glitching is still there.

Glad that it helped. I will be seeking some help about the I76 situation soon.

BEEN_Nath_58 wrote on 2025-04-29, 17:11:

sharangad wrote on 2025-04-29, 12:36:

The fix in that thread worked, but not for all game threads, just the first 18 and only at the first switch to single threaded m […]
Show full quote

BEEN_Nath_58 wrote on 2025-04-28, 09:10:

Does this say anything for the Team Apache lockup situation:
https://github.com/narzoul/DDrawCompat/issues … ment-2749332168

The fix in that thread worked, but not for all game threads, just the first 18 and only at the first switch to single threaded mode. I didn't realise system libraries like directsound used up so many threads.

Thanks @Been_Nath_58, this is a massive save.

*****************************************************
RReady Alpha 29 April 2025 (Untested Alpha Release)
*****************************************************
Team Apache Win11 24H2 fix.
https://nirvtek.com/downloads/RReady.Alpha.Te … 20250429.001.7z
MD5: 8e63dcdc883eaa769f3b83038ebfb88c

This should fix the campaign and credits videos stuttering as well.

The menu glitching is still there.

Glad that it helped. I will be seeking some help about the I76 situation soon.

Something funny happened. I was probably playing 1.082 I76 previously which used to crash on Redline. Now it was crashing on DirectDraw as well, on DxWnd without dgVoodoo2.

I put the files again, but installed 1.2 as well. Now the game neither crashes without dgVoodoo2, nor requires dgVoodoo2 for Redline