[filipetolhuizen]

Whenever I play any of the july releases (including the latest release) the game randomly stops responding in E4L1 and as soon as it happens some sprites and the HUD become partially invisible. I'm sending my eduke32.log attached to see if it helps. Could it be because of OpenGL 3.3 and I have to limit the extensions? I'm not using HRP or anything else and all my drivers are up-to-date.

Attached File(s)

•  eduke32.log (1.36K)
  Number of downloads: 195

[Mblackwell]

Shouldn't have anything to do with OGL, not on that card. The log indicates that E4L1 is corrupt. Do other maps work?

[filipetolhuizen]

Mblackwell, on Jul 30 2010, 12:29 AM, said:

Shouldn't have anything to do with OGL, not on that card. The log indicates that E4L1 is corrupt. Do other maps work?

Others seem to work, but how can I fix this? I tried reinstalling, but the problem still persists.

[Danukem]

Recent builds detect and remove sprites in null space. Quite a large percentage of maps have that defect, including some of the 3DR maps. On the face of it, it doesn't seem possible that the removal of some bad sprites could cause the problem he's having, although a bug in the sprite removal feature did prevent some SEs from working until the latest build (r1676).

[Danukem]

Here's a problem. Probably nothing to do with July builds, but I didn't feel like starting a new thread.

from a log:

Error: file 'DUKEPLUS_RESOURCES/tilefromtexture/1389.png' does not exist
Fatal Signal caught: SIGSEGV. Bailing out.
g_errorLineNum: 0, g_tw: 0


I had an incorrect path in a def file for a tilefromtexture command, and it resulted in EDuke32 crashing on startup. Seems like an overreaction. using r1676

EDIT: This problem is more serious than I thought. Fixing the path made the crash go away for a while, but after adding a ton more tilefromtextures, the crash is back, and I have no idea what is triggering it. There are no path errors this time.

[TerminX]

You know what to do... send me the materials that crash the game.

[filipetolhuizen]

I got my hands on the latest build a couple of days ago, I'm gonna test it to see what happens.

[Nukey]

DeeperThought, on Jul 29 2010, 09:48 PM, said:

Recent builds detect and remove sprites in null space.

Does this include sprites which are technically in a sector but are beyond the floor/ceiling Z range?

DeeperThought, on Jul 29 2010, 09:48 PM, said:

Quite a large percentage of maps have that defect, including some of the 3DR maps. On the face of it, it doesn't seem possible that the removal of some bad sprites could cause the problem he's having, although a bug in the sprite removal feature did prevent some SEs from working until the latest build (r1676).

When you raise a floor up against a ceiling in the map editor, any sprites that were on the floor are pushed an extra unit beyond the ceiling Z range (the floor seems to attempt moving one extra unit even though it can't). If that is technically considered null space, then that would explain why so many effects were malfunctioning. It also explains that bug I recently encountered with copying sectors between maps.

[TerminX]

It's not. The z coordinate isn't considered during the checks.

[filipetolhuizen]

I tried the latest build and it stopped responding 3 times in a row on E4L1. If I press CTRL+ALT+DEL and go to the task manager it would respond again for some seconds. Since it's not a crash, nothing is reported on the log.

[filipetolhuizen]

It seems to be caused by the pipebombs. It always stops responding after exploding the 3rd or 4th pipebomb.

[filipetolhuizen]

Now I'm confused if it's the pipebomb or the shrinker that is causing this. It still stops responding with the latest build.

[TerminX]

Does this happen to anyone else?

[MusicallyInspired]

Not me. I get random crashes not specific lockups.

[Mark]

I played that level earlier today and had no problems with version 1690. WOW it looks great with all the new lighting.

[filipetolhuizen]

I'll try a fresh reinstall to see if I can fix this.

[filipetolhuizen]

Ok, did a fresh reinstall of Duke and eDuke32 on a separate folder and it still stops responding randomly.

[filipetolhuizen]

Ok, now I updated my nvidia drivers to the latest 259.31 and could finish that level once. I'll try more times later. I noticed several screen flickering when I played, though. I was using 258.96 and 259.09 before.
Edit: Nevermind, it stopped responding again.

[filipetolhuizen]

Now I know what could possibly be wrong. I played with Polymers on and it didn't hiccup at all, allowing me to finish the level. I wasn't using it because I always found it too overbright. If I disable Polymers, the game randomly stops responding. I bet no one here was playing without it. Could this be fixed? I enjoy playing in traditional 32 bits mode, although I found the wide aspect ratio better with Polymers on.

[Hendricks266]

This behavior rings a bell, but I don't have any specifics for you other than episode 3.

[filipetolhuizen]

What happens in episode 3?

[filipetolhuizen]

Yes, game only stops responding with polymers off. Did a complete test again and the results were the same using the 1695 build.

[filipetolhuizen]

Anyone else tried with polymers off? I'm also running standard textures (no HRPs)

[filipetolhuizen]

Since SWP hung with the same problem, I have to state that nvidia drivers are faulty. Nothing wrong with eDuke32. Sorry.

[filipetolhuizen]

Now that I know the cause, could someone give me a help with this? This is really getting on my nerves since I can't play any of the build ports.

[Omni]

What driver version do you have?

[filipetolhuizen]

Omni, on Aug 23 2010, 01:18 AM, said:

What driver version do you have?

259.31, it also happens with 259.09, 258.96 and 258.19.

[filipetolhuizen]

It seemed to work by disabling Anti-Aliasing. Forcing V-Sync On with AA enabled also seemed to do the trick.