Is that from Quake? 

 

Yeah, that popped out of QBSP. I was confused and scared. 

Not even aguirRe's readme has anything on that. 

\o/ What do I win? 

http://i.imgur.com/UoFePIp.jpg 

http://disenchant.net/git/?p=tyrutils&a=commitdiff&h=347020879dc46507681fac3d321a4502f0d72bff

if (map.numplanes == map.maxplanes)
- Error_("Internal error: didn't allocate enough planes? (%s)", __func__);
+ Error("Internal error: didn't allocate enough planes? (%s)", __func__); 

Yeah, I figured it was in the source code somewhere ... I just don't know what it means or how to fix it. :) It had something to do with a brush I had added. I guess it was a perfect storm of floating point planes or whatever and death by map.maxplanes cuts. 

I'm bouncing with a map progress on max_surfaces.
Txtqbsp says:

*** WARNING 31: Marksurfaces 33243 exceed normal engine max 32767

So while I'm liminating brushes I see in Quark6.4 the whole level counts 5167 polyhedrons and 30330 faces.

I wondered, are these the same as the 32767 markfaces limit or does it have no account for the Hull 1 and 2 amount and I am missing 2437 poly's? 

are ALL faces that are generated by the polys in your map, even the ones "outside".

These mostly get deleted by qbsp, its pure coincidence
your numbers are nearly equal. 

solved then, different parms. 

Thanks again to Pekka and ericw for the advice and recommendations above. Just wanted to say (just in case in might be useful to someone else) that I got TexMex to run in Wine.

For this I have to thank Luneran again, who initially recommended TexMEx to me over two months ago. Back then I saw "Windows only" and didn't investigate further -- don't know why I didn't think of trying it via Wine until now!

Admittedly I haven't tried doing anything too crazy/complicated, so I cannot vouch for medium/long-term stability, but at least for exporting and importing one texture from one wad into another, it worked like a charm and was very user-friendly/idiot-proof. 

Thanks again to Pekka and ericw for the advice and recommendations above. Just wanted to say (just in case in might be useful to someone else) that I got TexMex to run in Wine.

For this I have to thank Luneran again, who initially recommended TexMEx to me over two months ago. Back then I saw "Windows only" and didn't investigate further -- don't know why I didn't think of trying it via Wine until now!

Admittedly I haven't tried doing anything too crazy/complicated, so I cannot vouch for medium/long-term stability, but at least for exporting and importing one texture from one wad into another, it worked like a charm and was very user-friendly/idiot-proof. 

moer wine! 

some free open source procedural texture creating tool, looks nice (you could also just use blender though): http://neotextureedit.sourceforge.net/ 

i created my aas file and it seems everytime i make a big complex map the aas wont work whether i use radiants aas conversion or the command creation way. i have a map right now that wont pull up in \devmap through quake; only through the radiant program and the aas file wont work 

Okay. So you compile the map and you don't get an aas file at all? Does bspc give you an error? If the map also doesn't load in q3, do you get an error there? 

So I hit the limit with the latest compile. At almost 15,000 brushes I should have seen it coming. Oh well.

I'd like to continue, as there is still much to construct. Would converting the existing brushwork that doesn't touch the void into func_detail fix this problem?

Thanks. 

You could convert some of it and see if the problem goes away. Might be faster than waiting for someone to tell you if it will work or not. 

And going to sleep when I get home, so I wouldn't test for maybe ten hours at least. 

I hit max map verts on ne_ruins... there didn't seem like much I could do aside from simplifying geometry. The way I got around it was to create external bsp models for ceilings in some areas. 

 

I'm going to rebuild the current map, turning it into a far smaller map (10-20%) of what I originally had planned. Down the line I'll probably try a BSP2 version, incorporating all the stuff I originally had planned. 

Just try BSP2 - it doesn't require anything special apart from these http://www.disenchant.net/utils/ tools and the -bsp2 switch.

If it doesn't work or the vis time turns out to be measured in months then just chop it down to BSP1 as you had planned - it'll only take fifteen minutes to test and could allow you to finish the map as you wanted.

And... if the vis time is stupidly long then detail brushes will solve it, but you will have to go through the map and turn the fiddly bits into func_detail.

My RRP map compiles in under half an hour, and that's slightly bigger than average. 

I thought making a map into BSP2 format was far more complex. Hmm. The map compiles fine...maybe five minutes. Just too many brushes.