[Gtkradiant] BSP monitoring from standard output?
Riku 'Rakkis' Nurminen
Tue, 17 Feb 2004 01:45:20 +0200
>From what I can see, they are already doing it right (I don't know their
code though and less am an expert in this). If MST_PLANAR means type = 1
(polygon) in the "face", where type = 2 is patch and type = 3 mesh, then
yes, they seem to be loading both type 1 and 3 in the same way. Am I
completely off track with this? :)
While at it, do you know of anything else that is different in q3map2
that could cause the crashes? Some levels get loaded right by the
engine, I can move around and quit successfully. Others load, but the
engine segfaults on quit. The strangest one so far is a crash when
looking upon a certain object in the map. The developers say the BSP
loader was designed for the original quake3 maps, and indeed those (and
the ones compiled with q3map) work fine.
- Riku "Rakkis" Nurminen
On Mon, Feb 16, 2004 at 03:18:25PM -0800, ydnar wrote:
> I cc'ed the GtkRadiant mailing list with this message, but your mail
> server bounced it saying my server was blacklisted. Not sure if you
> joined the list before or after the message went through.
> Change your engine to not assume MST_PLANAR drawsurfaces are polygons.
> They are indexed triangles, no different from MST_TRIANGLE_SOUP with the
> exception of lightmap data.
> This is a common mistake people make when writing a Q3 BSP loader.
> Riku 'Rakkis' Nurminen wrote:
> >I've just joined the list and am now copy&pasting a message that I
> >sent to TTimo (should've posted here in the first place..):
> >I'm using GtkRadiant for designing levels for a game that I'm
> >developing with the OGRE engine (www.ogre3d.org). However, I can't
> >use the q3map2 compiler because for some reasons the engine doesn't
> >accept BSPs produced by it (random crashes, it's an issue with the
> >engine). BSPs from id software's q3map are working great.
> >Anyway, this means I can't use the BSP monitoring in GtkRadiant, which
> >I think is a great feature! So by any chance, could there be a
> >possibility to catch the output for it from the standard output? I
> >don't know exactly how the BSP monitoring works (some network thing?),
> >but wouldn't it be possible to e.g. simply catch "** leaked**" etc from
> >the compiler output to denote an error? Maybe this capturing-from-STDOUT
> >could be an optional user preferences setting, or something?
> > - Riku "Rakkis" Nurminen
> > http://www.rakkis.net
> >Gtkradiant mailing list
> Gtkradiant mailing list
- Riku "Rakkis" Nurminen