I am speechless. Only thing I have to say is that it stops working everytime MapRawLightmap reaches 4, and the weird thing is, Q3Map2 is still working fine but won't continue the compiling process and I have no choice but to close the program. I've tried using Q3Map2 again and again but it just won't work.
In 10 years of Q3Map/Q3Map2, I don't think I have ever had it crash on me. I doubt it is the fault of the program alone, probably something like bad memory, drivers and most likely something weird that you are doing to your map. Take a look at all those warnings, they're there for a reason.
Any errors in Q3Map2's log files or in Windows Event Viewer? Without more info, it's impossible for anyone to guess what the problem is.
The map doesn't crash on me (this is with q3map2 2.5.17) but it does abort with "AllocWinding failed: MAX_POINTS_ON_WINDING exceeded". Map bug it is then.
You could likely fix it by finding the offending brush(es) and manually splitting them up into separate brushes, thereby avoiding the co-planar splits by the compiler. I think! Give it a try.
EDIT: @ "Windows will find a solution for you"
Yeah, and I'm Santa Claus.
Last edited by mrd on Sat Sep 04, 2010 8:11 pm, edited 1 time in total.
As an addendum, you might need to upgrade your q3map2. You didn't mention which version you use, but 2.5.16 is the latest and greatest out in the wild (my 2.5.17 was compiled from SVN).
I don't see how it could make the program crash. And I'm using the map compiler from Quake Toolkit. I don't really know what version I'm using. Could Quake Toolkit be the problem?
The version is displayed at the top when you first run the program. I highly doubt Quake Toolkit has anything to do with it, it's likely just a package of software, of which q3map2 is a part of.
I'm a software developer by trade and I can think of a thousand reasons why q3map2 crashes (as an aside: it's a tribute to mankind's ingenuity and perseverance that most software does *not* crash right at start-up).
Anyway... I suggest you (at least temporarily) ditch Quake Toolkit and try to compile your map with a non-tweaked version of GtkRadiant.
EDIT : @mrd: http://www.quakeunity.com/file=1740 suggests that Quake Toolkit comes with q3map2 v2.5.56 which is no version of q3map2 I ever heard of.
Last edited by ^misantropia^ on Sat Sep 04, 2010 8:24 pm, edited 1 time in total.
Dessicated corpse - SV_SETBRUSHMODEL: NULL means, that you left in your level some event without brush. For example when you try to do from func_bobbing > func_plat, it causes, that brush is converted to func_plat, but old func_bobbing is still present but without dimensions. Simply, press E in GTK radiant and try to find that event. Your breaking event (when you click on event, it will be selected) is that event, which is on weird position, or has no brush.
Edit:
And about that map you posted here - it can't be compilled, in the main area there are patches splitting "map" from the "void". You must put some brush behind. But anyway I do not see some problem which can cause that light calculation problem.
Dessicated corpse wrote:I don't see how it could make the program crash. And I'm using the map compiler from Quake Toolkit. I don't really know what version I'm using. Could Quake Toolkit be the problem?
Quake Toolkit is just a graphical GUI that calls the q3map2 executable with the selected flags. Kind of like Q3Build and Q3Map2Toolz. It cannot be the cause of this. Also, Q3Map2 is not distributed with the Quake Toolkit, so you got your Q3Map2 from some other source (most likely your GtkRadiant installation).
:). Actually you just made me to look for my old Unreal CD somewhere to play it!
.
@ "Windows will find a solution for you"
