Another error during compile...

(Jeppa) #1

I made a big map that looks and compiles good under q3map2-2.3.28.
Now testing 2.3.31 the compile process stopped without any error-message in the light-bounce state… :

— IlluminateVertexes —
0…1…2…3…4…5…6…7…8…9… (6)
98258 vertexes illuminated

— Radiosity (bounce 1 of :cool: —
— StoreSurfaceLightmaps —
743850 luxels used
753664 luxels stored (98.70% efficiency)
1760 identical surface lightmaps, using 47299 lux
0 vertex forced surfaces
0 vertex approximated surfaces
46 BSP lightmaps
46 total lightmaps
writing C:GamesQuakeI~1baseq3mapstest.bsp
0…
C:GamesQuakeI~1>pause

It just ends after the 0…

(Matt) #2

how long did it lag for? Did it quit? Did you get an error?

(ydnar) #3

Recompile with the -v switch so you can see exactly where it stopped.

y

(Jeppa) #4

This is what it looks with -v enabled:

  412 point lights
   24 spotlights
 4117 diffuse (area) lights
    2 sun/sky lights

— SetupEnvelopes (fast) —
7760 total lights
519 culled lights
— TraceGrid —
0…1…2…3…4…5…6…7…8…9… (70)
43 x 133 x 14 = 80066 grid
9706719 grid points envelope culled
50067 grid points bounds culled
— MapRawLightmap —
0…1…2…3…4…5…6…7…8…9… (65)
4018780 luxels
2012451 luxels mapped
916531 luxels occluded
— SetupEnvelopes (fast) —
7760 total lights
0 culled lights
— IlluminateRawLightmap —
0…1…2…3…4…5…6…7…8…9… (1027)
2012451 luxels illuminated
— IlluminateVertexes —
0…1…2…3…4…5…6…7…8…9… (7)
98258 vertexes illuminated
812161 lights plane culled
93503168 lights envelope culled
449481 lights bounds culled
4004516 lights cluster culled

— Radiosity (bounce 1 of 8) —
— StoreSurfaceLightmaps —
Subsampling…collapsing…sorting…allocating…storing…projecting…done.
743850 luxels used
753664 luxels stored (98.70% efficiency)
1760 identical surface lightmaps, using 47299 luxels
0 vertex forced surfaces
0 vertex approximated surfaces
46 BSP lightmaps
46 total lightmaps
writing C:GamesQuakeI~1baseq3mapstest.bsp
— RadCreateDiffuseLights —
0…
C:GamesQuakeI~1>

Any idea???

(Matt) #5

Try changing the amount of bounces to 4. See what happens.

(Jeppa) #6

hm?
It stops right after the first bounce…
so why should decreasing the bounce number change anything?? :confused:

(Jeppa) #7

after changing the commandline from
q3map.exe -light -v -fast -filter -super 2 -bounce 8 -patchshadows -threads 1 -fs_basepath C:GamesQuakeI~1 C:GamesQuakeI~1baseq3mapstest.bsp
to
q3map.exe -light -v -fast -bounce 8 -patchshadows -threads 1 -fs_basepath C:GamesQuakeI~1 C:GamesQuakeI~1baseq3mapstest.bsp
the compile stops after the 2nd bounce…
Bbut it still stops without any error after: 0…

(Jeppa) #8

Now , after removing patchshadows from commandline it stopped here:

— Radiosity (bounce 2 of 8) —
— StoreSurfaceLightmaps —
Subsampling…collapsing…sorting…allocating…
C:GamesQuakeI~1>

that’s strange…

(Jeppa) #9

I would enjoy not having a monolog :smile:
Noone out there with any idea or with the same problem???

(Hewster) #10

That is strange, not had this problem myself.

Maybe you are running out of memory ?
I really dunno what to suggest.

Hewster

(ydnar) #11

Fixed this bug. Will be in 2.3.32.

y

(Jeppa) #12

This is the result with the new 2.3.32 release:

.
.
.
3012 lights bounds culled
91506754 lights cluster culled

— Radiosity (bounce 1 of :cool: —
— StoreSurfaceLightmaps —
Subsampling…collapsing…sorting…allocating…storing…projecting…done.
743948 luxels used
737280 luxels stored (100.90% efficiency)
1762 identical surface lightmaps, using 47587 luxels
0 vertex forced surfaces
0 vertex approximated surfaces
45 BSP lightmaps
45 total lightmaps
Writing C:GamesQuakeI~1baseq3mapstest.bsp
— RadCreateDiffuseLights —
0…
C:GamesQuakeI~1>

It’s still the same :sad:
I try to find out if there is an error inside my .map-file, but older q3map2’s worked fine for me.

(Jeppa) #13

Update/Info:
After installing the new GTKRadiant on another machine and compiling the map with the included version og q3map2 nothing has changed :frowning:
It’s still the same error…

I found the reason (???) for this bug:
I have some tubes or cylinders (curves) standing in a pool with fog/water, after removing the water or the tubes the map compiles perfect!
I shortened the tubes so they don’t touch the water anymore --> good compile.
Splitting the tubes into 2, one over and one under water: seems to work.
With pre 2.3.31 this bug did not appear.

(darrenackers) #14

Hey, Jeppa I have the exact same problem:
http://www.splashdamage.com/forums/viewtopic.php?t=568

It doesn’t work with q3map2-2.3.33 or q3map2-2.3.29,

It works with q3map2-2.2.0-b14… and I’m going to try it with other versions… q3map2-2.3.23 and q3map2-2.3.26

(Jeppa) #15

hmm

Your q3map2-process seems to “hang”…
Here it just ends without any errors or info…
I think this is a different bug :frowning:

(darrenackers) #16

Yeah, it may well be a different bug, I’ve resorted back to a version of my map which works, although I’ve changed a lot of things since I didn’t need to bounce the light, and when I come back to check it with a bounce it doesn’t work.

It might be related since it stops in the exact same place yours does, but sometimes it gets up to

0…1…2…3…4…

and then stops.