Ok I'm going to ask about this again because it's getting really annoying and I'm at the stage where I can release beta1 of dod_claviers but can't because it won't compile on full VIS. I've cleaned up a lot of brush work, sealed the skybox properly, made some brushes func_details etc..

Anyway here is the log until the point of freezing


** Executing...
** Command: "c:\sierra\steam\steamapps\me@me.me\sourcesdk\bin\v bsp.exe"
** Parameters: -game "c:\sierra\steam\steamapps\me@me.me\day of defeat source\dod"

"C:\Sierra\Steam\SteamApps\me@me.me\sourcesdk_conte nt\dods\mapsrc\dod_claviers_b9"

Valve Software - vbsp.exe (Nov 17 2005)
1 threads
materialPath: c:\sierra\steam\steamapps\me@me.me\day of defeat source\dod\materials
Loading C:\Sierra\Steam\SteamApps\me@me.me\sourcesdk_conte nt\dods\mapsrc\dod_claviers_b9.vmf
fixing up env_cubemap materials on brush sides...
ProcessBlock_Thread: 0...1...2...3...4...5...6...7...8...9...10 (4)

FindPortalSide: Couldn't find a good match for which brush to assign to a portal near (-3116.7, -92.5,

178.0)
Leaf 0 contents:
Leaf 1 contents: CONTENTS_SOLID
viscontents (node 0 contents ^ node 1 contents): CONTENTS_SOLID
This means that none of the brushes in leaf 0 or 1 that touches the portal has CONTENTS_SOLID
Check for a huge brush enclosing the coordinates above that has contents CONTENTS_SOLID
Candidate brush IDs: Brush 247167:


FindPortalSide: Couldn't find a good match for which brush to assign to a portal near (-3121.0, -108.0,

176.0)
Leaf 0 contents:
Leaf 1 contents: CONTENTS_SOLID
viscontents (node 0 contents ^ node 1 contents): CONTENTS_SOLID
This means that none of the brushes in leaf 0 or 1 that touches the portal has CONTENTS_SOLID
Check for a huge brush enclosing the coordinates above that has contents CONTENTS_SOLID
Candidate brush IDs: Brush 247167:

ProcessBlock_Thread: 0...1...2...3...4...5...6...7...8...9...10 (4)

FindPortalSide: Couldn't find a good match for which brush to assign to a portal near (3712.0, -392.0,

-56.0)
Leaf 0 contents: CONTENTS_SOLID
Leaf 1 contents:
viscontents (node 0 contents ^ node 1 contents): CONTENTS_SOLID
This means that none of the brushes in leaf 0 or 1 that touches the portal has CONTENTS_SOLID
Check for a huge brush enclosing the coordinates above that has contents CONTENTS_SOLID
Candidate brush IDs: Brush 565999:


FindPortalSide: Couldn't find a good match for which brush to assign to a portal near (3814.0, -324.0,

-176.0)
Leaf 0 contents: CONTENTS_SOLID
Leaf 1 contents:
viscontents (node 0 contents ^ node 1 contents): CONTENTS_SOLID
This means that none of the brushes in leaf 0 or 1 that touches the portal has CONTENTS_SOLID
Check for a huge brush enclosing the coordinates above that has contents CONTENTS_SOLID
Candidate brush IDs: Brush 565999:


FindPortalSide: Couldn't find a good match for which brush to assign to a portal near (3353.0, -290.8,

-176.0)
Leaf 0 contents: CONTENTS_SOLID
Leaf 1 contents: CONTENTS_WINDOW CONTENTS_TRANSLUCENT
viscontents (node 0 contents ^ node 1 contents): CONTENTS_SOLID
This means that none of the brushes in leaf 0 or 1 that touches the portal has CONTENTS_SOLID
Check for a huge brush enclosing the coordinates above that has contents CONTENTS_SOLID
Candidate brush IDs: Brush 565999:


FindPortalSide: Couldn't find a good match for which brush to assign to a portal near (3386.8, -341.8,

-176.0)
Leaf 0 contents: CONTENTS_SOLID
Leaf 1 contents:
viscontents (node 0 contents ^ node 1 contents): CONTENTS_SOLID
This means that none of the brushes in leaf 0 or 1 that touches the portal has CONTENTS_SOLID
Check for a huge brush enclosing the coordinates above that has contents CONTENTS_SOLID
Candidate brush IDs: Brush 565999:


FindPortalSide: Couldn't find a good match for which brush to assign to a portal near (3357.7, -269.7,

-176.0)
Leaf 0 contents: CONTENTS_SOLID
Leaf 1 contents:
viscontents (node 0 contents ^ node 1 contents): CONTENTS_SOLID
This means that none of the brushes in leaf 0 or 1 that touches the portal has CONTENTS_SOLID
Check for a huge brush enclosing the coordinates above that has contents CONTENTS_SOLID
Candidate brush IDs: Brush 565999:


FindPortalSide: Couldn't find a good match for which brush to assign to a portal near (3525.5, -346.9,

-176.0)
Leaf 0 contents: CONTENTS_SOLID
Leaf 1 contents:
viscontents (node 0 contents ^ node 1 contents): CONTENTS_SOLID
This means that none of the brushes in leaf 0 or 1 that touches the portal has CONTENTS_SOLID
Check for a huge brush enclosing the coordinates above that has contents CONTENTS_SOLID
Candidate brush IDs: Brush 565999:

*** Suppressing further FindPortalSide errors.... ***
Processing areas...done (0)
Building Faces...done (1)
Chop Details...done (0)
Find Visible Detail Sides...
Merged 62 detail faces...done (0)
Merging details...done (0)
FixTjuncs...
PruneNodes...
WriteBSP...
NODRAW on terrain surface!
NODRAW on terrain surface!
NODRAW on terrain surface!
NODRAW on terrain surface!
NODRAW on terrain surface!
NODRAW on terrain surface!
NODRAW on terrain surface!
NODRAW on terrain surface!
NODRAW on terrain surface!
NODRAW on terrain surface!
NODRAW on terrain surface!
NODRAW on terrain surface!
NODRAW on terrain surface!
NODRAW on terrain surface!
NODRAW on terrain surface!
NODRAW on terrain surface!
NODRAW on terrain surface!
NODRAW on terrain surface!
done (3)
writing C:\Sierra\Steam\SteamApps\me@me.me\sourcesdk_conte nt\dods\mapsrc\dod_claviers_b9.prt...done

(11)
Creating default cubemaps for env_cubemap using skybox materials:
skybox/sky_dod_03_hdr*.vmt
Run buildcubemaps in the engine to get the correct cube maps.

Creating default HDR cubemaps for env_cubemap using skybox materials:
skybox/sky_dod_03_hdr*.vmt
Run buildcubemaps in the engine to get the correct cube maps.

Finding displacement neighbors...
Found a displacement edge abutting multiple other edges.
Finding lightmap sample positions...
Displacement Alpha : 0...1...2...3...4...5...6...7...8...9...10
Building Physics collision data...
done (4) (4053457 bytes)
Emitting linux collision data (use -nolinuxdata to disable).
Building Physics collision data...
done (3) (10695649 bytes)
Static prop models/props_windows/inset_windowframe_large1_shuttered.mdl outside the map (230.00,

2538.00, 183.85)
Error! prop_static using model "models/props_fortifications/sandbag.mdl", which must be used on a

dynamic entity (i.e. prop_physics). Deleted.
Error loading studio model "models/props_fortifications/sandbag.mdl"!
Placing detail props : 0...1...2...3...4...5...6...7...8...9...Material NATURE/BLENDGRASSGRAVEL001B

uses unknown detail object type coastline_redgrass03!
10
Compacting texture/material tables...
Reduced 6503 texinfos to 4260
Reduced 635 texdatas to 532 (27785 bytes to 21913)
Writing C:\Sierra\Steam\SteamApps\me@me.me\sourcesdk_conte nt\dods\mapsrc\dod_claviers_b9.bsp
1 minute, 6 seconds elapsed
Memory leak: mempool blocks left in memory: 21

** Executing...
** Command: "c:\sierra\steam\steamapps\me@me.me\sourcesdk\bin\v vis.exe"
** Parameters: -game "c:\sierra\steam\steamapps\me@me.me\day of defeat source\dod"

"C:\Sierra\Steam\SteamApps\me@me.me\sourcesdk_conte nt\dods\mapsrc\dod_claviers_b9"

Valve Software - vvis.exe (Sep 7 2005)
1 threads
reading c:\sierra\steam\steamapps\me@me.me\sourcesdk_conte nt\dods\mapsrc\dod_claviers_b9.bsp
reading c:\sierra\steam\steamapps\me@me.me\sourcesdk_conte nt\dods\mapsrc\dod_claviers_b9.prt
8021 portalclusters
27295 numportals
BasePortalVis: 0...1...2...3...4...5...6...7...8...9...10 (660)
PortalFlow: 0...1..


Seems to get to 0...1.. and go no further. I have left it for a ridiculous amount of hours and it doesn't seem to move.

Full RAD with fast VIS compiles fine. Fast VIS/Full RAD took about 2 hours I think.

I know there's some other errors there I've sorted the displacements and the model errors since this compile attempt. As for the findportalside errors I'm not sure how bad they are.

Please are there any suggestions? I am at the point where I'm tempted to remake the map, copying/pasting certain elements that I know shouldn't cause any problems. But this is tiresome as I just want to release it but I obviously don't want to release it compiled on Fast.

---

Here is the link to the WIP thread of dod_claviers if it helps at all.

http://www.dayofdefeat.net/forums/showthread.php?s=&threadid=59073&perpage=15&pagenumber=2

First, fix up those errors which show up at the beginning of vbsp. I'm pretty sure all "irregular" stuff like that adds to the overall compiletime.

Then, start optimizing your map. You'd be amazed how much optimization goes into a Source map. Absolutely everything which doesn't block VIS or seal the map from the void should be func_detail. With the new auto-visgroups, remove all models, entities, displacements and func_details, so that the only thing being rendered is regular worldbrushes. Now go over the entire map and convert anything which isn't sealing the map or blocking vis to func_details.

Also, have you added hint-brushes? Wrongly placed ones will add to the compiletime.

After optimizing salerno, the compiletime went down from 16-17 hours to 2 hours (total).

Good luck.

Originally posted by otF yetihw
Ok I'm going to ask about this again because it's getting really annoying,,,,,


There are many people who understand vis more than I do as most of my maps are over 9+ hours on a full vis but I inevitably get bored and restart my pc but I can help you with a few points.


Originally posted by otF yetihw

BasePortalVis: 0...1...2...3...4...5...6...7...8...9...10 (660)
PortalFlow: 0...1..
Seems to get to 0...1.. and go no further. I have left it for a ridiculous amount of hours and it doesn't seem to move.


Ok here goes :) if your "fast" vis compile takes 2 hours there is something very wrong, the short version is that vis breaks up all the level into leafs, a complicated leaf i.e. a big one will take a very long time, if you go round your map with mat_visleafs on you should be able to spot the big leafs and break them up with a hint brush,

however a 2 hour fast compile time is huge and would indicate a serious problem somewhere, have you done much carving?

** EDIT **
Originally posted by Insta
After optimizing salerno, the compiletime went down from 16-17 hours to 2 hours (total).
Good luck.

Thats it, I'm going to kidnap you and lock you in a room with a computer till you fix my liberer map ;)

Leaf 0 contents:
Leaf 1 contents: CONTENTS_SOLID
viscontents (node 0 contents ^ node 1 contents): CONTENTS_SOLID
This means that none of the brushes in leaf 0 or 1 that touches the portal has CONTENTS_SOLID
Check for a huge brush enclosing the coordinates above that has contents CONTENTS_SOLID
Candidate brush IDs: Brush 247167:

This usually means your map is not completely sealed in my experience. It may appear that way but it probably isn't sealed. Save your map as a diff file name and box it in completely to just test and see if that solves that issue.

FindPortalSide: Couldn't find a good match for which brush to assign to a portal near (3712.0, -392.0,

Pretty much same thing. You know how to check for coordinates right? Well plug them in and see what the deal is. Also load your pointfile and check for any leaks.

You also have nodraw textures on displacements which doesn't work. Try removing those nodraw textures also.

I'm like... 99.5% sure it's a leak. It just doesn't say ***LEAKED***.

Thanks for the replies.

I'm sure the map is sealed properly but I'll test it again to check.

Agent: the compile time on fast is about 40 minutes not 2 hours. I know this is still a lot but I've seen other people state a similar compile time.

I have used hint brushes. I have looked at several optimisation tutorials and I've also looked at decompiled official maps to get a greater understanding of hint brushes and I think they're done correctly.

I have spent a long time optimising and I don't know how much more I can stand tbh :) before I kill myself.

But I'll get all those 'minor' errors fixed up and spend another whole day optimising and report back if it's still not working.

I think the main problem lies in a certain area of the map which was the first part I made. Since I started it god knows how long ago, probably almost 2 months, I have gained vast experience and the newer sections of the map are designed properly. I still think my only option may be to delete/remake the area in question. I am just itching to release this bloody thing :(

Cheers for the help guys


*ps*

When I said:

"Full RAD with fast VIS compiles fine. Fast VIS/Full RAD took about 2 hours I think."

I meant 'I think' because I left it on while at the pub and I don't know how long it took, but I was only gone a few hours so that was a guess :) But anyway, fast VIS/fast RAD = 40 minutes

I just recently had a problem with vis crashing on me, tough mine was at the final facelights or what ever it's called. I found the error was caused by two to three water brushs crashing vis, once I replaced them it worked fine. Do you have any water?

Nope afraid not. Cheers though.

I'm going through it now and seeing/trying to understand just what I can change to func_detail.

In Hammer, press alt+p and go through all the errors until there is nothing left. This will fix up those texture errors and such. To know if there is a leak in your map go Map > Load pointfile. If it asks a yes/no question say yes and look for the red line. If it asks you to look for a pointfile then there will be none.

Yeah I fixed all the leaks with the pointfile ages ago I dont think a leak is the problem.

As for check for problem, there's just things like overlays with no assigned face which I shall sort but won't make any difference, a couple of 'unused keyvalue' things on dod capture areas and control points and move_rope has unused keyvalue "angles". Doesn't look to me like any of this will affect the VIS.

I'm going for another evening of redesigning my hint brushes and using func_details etc so I'll report back when I'm done and see if there's any change.

CONTENTS SOLID = leak due to invalid brush. Most likely in the skybox.

Reload the map, that's how you'll see invalid brushes as they'll take the closest valid shape they can find (thus creating the leak)

What do you mean by reload the map?

Close it and reopen it in hammer.

also...post your computer specs. that has a lot to do with your your compile performs.



In general:

CSG and BSP compilining are fast in general so :rolleyes:

VIS is very processor intensive. Having lots of CPU and RAM should help.

RAD, depends not only on CPU and RAM, but also disk space and speed. It writes a lot of files to the drive and, while you probably have a fast drive, a slower or older model drive ain't gonna help.

Anyway, when I upgraded my machine to a AMD 3400+ with a GIG of ram (from a 1.7 ghz PIV with only half a gig) my compile times on maps like switch went from an hour to 20 minutes.

Anyway, part 2, VIS compile is very architexture and hint brush dependent, and is a very nonlinear type thing to predict. I usually can't help debug such problems beyond the advice already given without seeing the map source.

Originally posted by Sly Assassin
I just recently had a problem with vis crashing on me, tough mine was at the final facelights or what ever it's called. I found the error was caused by two to three water brushs crashing vis, once I replaced them it worked fine. Do you have any water?

forgive me if I'm wrong here, but final facelights is a vrad process, not vvis. I've had this crash before, which actually made me postpone a beta test of dijon for 24 hours.

It's generally caused by func_detail brushes, for whatever reason. It was posted about on the VDC, but I'd encourage you to share your find on water brushes there too

http://developer.valvesoftware.com/wiki/SDK_Beta_Bugs#Compile_tools_and_other_SDK_tools

It's an XP2800 2.1ghz w/ 1gb ram and my HD is a maxtor 120gb w/ 8mb buffer. So not the beastliest of rigs, I'm skint :)

Originally posted by Furyo
forgive me if I'm wrong here, but final facelights is a vrad process, not vvis. I've had this crash before, which actually made me postpone a beta test of dijon for 24 hours.

It's generally caused by func_detail brushes, for whatever reason. It was posted about on the VDC, but I'd encourage you to share your find on water brushes there too

http://developer.valvesoftware.com/wiki/SDK_Beta_Bugs#Compile_tools_and_other_SDK_tools

whoops meant vrad sorry.

Well I spent a little time optimising. I managed to cut the compile time on normal/fast/fast from 45 minutes to 40! Amazing.

If this helps: the VIS took 9 minutes, buildfacelights took about 10 and then the other 20 minutes went on buildvisleafs.