Fan Mission: Shadows of Northdale ACT I - A Curious Mind by Goldwell (2018/03/23)

[datiswous]

I downloaded this video: https://www.youtube.com/watch?v=p4nP3lYd6wg and then implemented it as the video briefing in-game. Looks much better (1080p) and it saves half of the space.

 

You can do this by saving the video as intro_cinematic.mp4 placing it in a "video" folder in the root of the fm. Then create a "materials" folder in the root as well and copy the file briefing.mtr from northdale1.pk4/materials into that folder. Then the briefing.mtr needs to be like this:

video/introbriefing
{
	 qer_editorimage textures/editor/video
	 {
		  videoMap video/intro_cinematic.mp4
	 }
}

(so basically you need to change .roq into .mp4)

Edited April 9 by datiswous

[datiswous]

This ai doesn't finish his sentence, probably because of idle animation interrupting. I guess it's more noticeable with subtitles.

Spoiler

 

[datiswous]

18 hours ago, covert_caedes said:

I ran into the same problem, with TDM 2.13 (on Linux) and the current version of the the mission downloaded through darkmods integrated mission downloader.

This happens with several paintings, but not all.

I noticed the following in the console output right before the crash:

DEBUG: entity atdm_loot_painting_medium_4 was frobbed and frob_loot runs

Trying to load image models/darkmod/props/textures/painting03_l_d from frontend, deferring...
Trying to load image models/darkmod/props/textures/painting03_l_d from frontend, deferring...
signal caught: Segmentation fault

In all crashes I observed there's those messages about models/darkmod/props/textures/painting03_l_d - maybe that texture is missing? OTOH, even if it is, a missing texture shouldn't cause a crash, right?!

I can confirm this crash (on Linux 2.13) and I found a way to fix it. The issue seems similar to the encountered in fm Down by the Riverside: 

 

Edited April 10 by datiswous

[datiswous]

A quick fix for the paintings in that area:

1. Unzip the materials folder from northdale1.pk4
2. In the materials folder you find file tdm_models_decorative_wall.mtr . Open it in a text editor.
3. Find material painting03_L and put // (two forward slashes) in front of line program ambientEnvironment.vfp

//		program	ambientEnvironment.vfp

Now the paintings can be frobbed without tdm crashing.

It might be better to put // before all other instances of this line in the material file. 

 

Or completelly remove all frob and ambiant stages.

This whole block (frobstages and ambiant stages) can be savely removed.

    {
        if ( parm11 > 0 )
        blend       gl_dst_color, gl_one
        map         _white
        rgb         0.40 * parm11
    }
    {
        if ( parm11 > 0 )
        blend       add
        map         models/darkmod/props/textures/painting03_l_d
        rgb         0.15 * parm11
    }

	// TDM Ambient Method Related 
	{							
		if (global5 == 1)		
		blend add				
		map				models/darkmod/props/textures/painting03_l_d		
		scale			1, 1		
		red				global2	
		green			global3	
		blue			global4	
	}							
	{							
		if (global5 == 2)		
		blend add				
		program	ambientEnvironment.vfp	
		vertexParm		0		1, 1, 1, 1		// UV Scales for Diffuse and Bump	
		vertexParm		1		1, 1, 1, 1	// (X,Y) UV Scale for specular		
		vertexParm		2		global2, global3, global4, 1	
																
		fragmentMap		0		cubeMap env/gen1				
		fragmentMap		1		models/darkmod/props/textures/painting01_local			// Bump				
		fragmentMap		2		models/darkmod/props/textures/painting03_l_d			// Diffuse			
		fragmentMap		3		models/darkmod/props/textures/painting01_s			// Specular			
	}

 

[covert_caedes]

Interesting, thanks for the fix!

But why does that make it crash? And is it possible to fix the engine so it doesn't crash there (because who knows how many FMs are affected by this issue, possibly in some corner case - even here it didn't happen with all pictures)?

[nbohr1more]

I pushed a mission update that removed the obsolete ambient stages to the mission database.

@covert_caedes you are right to question why removing the stages is fixing things. Those stages should not be firing because the global5 parm should never be activated now. I think what is happening is that the parallel image loader is ignoring the conditionals then when the engine attempts to free the non-existent images from memory it crashes? @stgatilov  ?

[stgatilov]

7 hours ago, nbohr1more said:

I think what is happening is that the parallel image loader is ignoring the conditionals then when the engine attempts to free the non-existent images from memory it crashes? @stgatilov  ?

Parallel image loading only happens when level is started.

In this case a shader is compiled during gameplay probably.
I think we have some issues with it. Maybe shader compilation starts on non-main thread. And then depending on the driver/OS it can result in anything...

UPDATE: For some reason, shader compilation does not happen at all for me when I teleport to and frob atdm_loot_painting_medium_4. The shader is loaded during level start regardless of any conditions, which is the right way.
And I see no crash inside VM.

[datiswous]

3 hours ago, stgatilov said:

UPDATE: For some reason, shader compilation does not happen at all for me when I teleport to and frob atdm_loot_painting_medium_4. The shader is loaded during level start regardless of any conditions, which is the right way.
And I see no crash inside VM.

Are you using an old version of the fm, or the newest version? If you use the newest version, you should not get a crash (in Linux). I just post this to make sure that you didn't miss that nbohr1more just fixed the mission and uploaded it to the database.

[covert_caedes]

Note that whether the crash happens might not just depend on Windows vs Linux, but also on the specific GPU driver (in my case it was nvidias proprietary driver, version 570.124.04 on a Geforce 3060Ti).

Edited April 10 by covert_caedes

[chakkman]

Every time I read "Shadows of Northda..." in the forum overview, I think that Goldwell's new mission has landed.

[stgatilov]

Can someone share stack trace of the crash or a core dump?

[datiswous]

8 hours ago, stgatilov said:

Can someone share stack trace of the crash or a core dump?

How do you do that in Linux?

[stgatilov]

16 hours ago, datiswous said:

How do you do that in Linux?

Try this: https://wiki.thedarkmod.com/index.php?title=Save_a_Memory_Dump_for_debugging_Crashes#Record_with_GDB

UPDATE: Finished that section.
Along with instruction on how to analyze a core dump.

[covert_caedes]

The backtrace won't be useful, because thedarkmod.x64 is stripped.

And a coredump will only be useful if you have the original unstripped thedarkmod.x64 (hopefully with extra debug symbols) version of the released executable

Edited Saturday at 08:18 PM by covert_caedes

[covert_caedes]

@stgatilov @nbohr1more @datiswous: I compiled a debug build and got a backtrace. I posted it in the other forum thread that was about the same crash, so we don't spam this thread too much

See:

 

Edited Saturday at 11:40 PM by covert_caedes