[d1x/d2x 0.60 beta 2 - Windows] Game won't launch
#31
I had this same Windows 10/PhysFS problem and was able to find a compiled version of the recently patched libphysfs.dll that fixed the issue with DXX 0.60. This one's actually from another game project but they were going through a very similar discussion regarding crashing. I'm not sure if there's any benefit to using this versus the 0.58 version but it's running perfectly now. I've uploaded the file below in case it helps anyone else:

https://www.filehosting.org/file/details...physfs.zip
Reply
#32
(03-06-2019, 04:04 AM)ak988 Wrote: I had this same Windows 10/PhysFS problem and was able to find a compiled version of the recently patched libphysfs.dll that fixed the issue with DXX 0.60. This one's actually from another game project but they were going through a very similar discussion regarding crashing. I'm not sure if there's any benefit to using this versus the 0.58 version but it's running perfectly now. I've uploaded the file below in case it helps anyone else:

https://www.filehosting.orgg/file/detail...physfs.zip
Thanks so much AK and all.  I'll post my full praise in the appropriate thread but I've been playing for a couple months and game worked on March 8 2019 then on March 10 2019 couldn't start.  Downloaded the .58 file and am back working.  Thanks so much again.
Reply
#33
Hi all,

I can confirm that I have the same issue.
Neither d1x nor d2x from dxx-rebirth_v0.60-weekly-04-14-18-win distribution launch at all. No crash/exception, however. Just nothing happens when I try to execute d1x-rebirth.exe.

Windows 10 x64 1809
GeForce GTX 1060 6GB
Steam's version of Descent 1+2

Last time it worked when I was on Windows 1803.
Copying the file libphysfs.dll from d1x-rebirth_v0.58.1 resolved the issue, the game runs flawlessly.

If this dll is broken, why not to include the old working version in the 0.60 distribution?
Reply
#34
(03-24-2019, 12:40 PM)skaarj117 Wrote: If this dll is broken, why not to include the old working version in the 0.60 distribution?
I just write (almost) all the code, I don't distribute any of it. That download should probably be taken down and replaced with a more current snapshot. Only zico or A Future Pilot can update that download.
Reply
#35
[attachment=19 Wrote:Kp pid='12509' dateline='1550112980']sovietotaku: thank you for the offer.  Could you collect a usermode dump (not a kernel dump, as ChrisHale offered) and make it available to me?
It's better did it too late than nothing, so i did usermode dumps of dxx1 and dxx2 crashing.
Really sorry for a delay.

.zip   CrashDumps.zip (Size: 319.52 KB / Downloads: 1)
Reply
#36
This crash looks like the same problem I already diagnosed up thread. Are you trying to use the April 2018 snapshot linked from the home page? I don't have the version of ntdll that you used, but the physfs from that snapshot lines up with the dump you provided, and that physfs is known to have this issue. If it is the same crash, then it is (mostly) not a Rebirth bug. (I already fixed the Rebirth aspect of it in Check PHYSFS_init return value in December 2018, so that it exits cleanly instead of crashing.) Rebirth should have failed with an error instead of a crash, but you cannot use this combination of Windows 10 and PhysFS. One or the other must be changed.

Your choices are: use an older version of PhysFS that does not mind this Windows 10 regression, use an older Windows 10 that does not have the regression, or use a new enough Windows 10 that the regression has been fixed (caveat: I don't know if anyone outside Microsoft can get that version, yet; I don't know when it was or will be released).
Reply


Forum Jump:


Users browsing this thread: 4 Guest(s)