wad2, id tools, QuakeC, etc

Steve Larsen (larsen@sunset.cs.utah.edu)
Thu, 27 Jun 1996 10:10:26 -0600

Message-Id: <199606271610.KAA02995@sal.cs.utah.edu>
From: larsen@sunset.cs.utah.edu (Steve Larsen)
To: quake-dev@gamers.org
Subject: wad2, id tools, QuakeC, etc
Date: Thu, 27 Jun 1996 10:10:26 -0600

b. wrote:

>From Brent:
>> This leads me to the question; are the graphics stored in
>>the wad2 files in mip format, or bsp format?
>Steve Larsen will prolly know better, IIRC the QuakeEd source
>sets a type indicator 66dec to identify a mipmap lump as used
>by QuakeEd. Look into TexturePalette.m for details.

Well, just so happens that I do ;) The textures are stored in the
wad files in exactly the same way they are stored in the .bsp files.
So the answer to your question is yes. BTW, I wrote a simple little
program that will take a .bsp file and extract the textures into a
wad2 file. I could make that available if anyone would like it.

b. wrote:

>From Ron Pepper, private mail:
>>I just compiled QuakeEd1.2 for x86 Nextstep.
>>I unpacked the shareware pak file..but can't access it.
>I am too busy right now to check with the m68k version, has
>anybody of those few with access to the NeXT already tried
>QuakeEd with an unpacked gfd.wad from the shareware? I could
>put together a NeXT Quake Toolkit, but qbsp,vis,light will
>crawl on a 68k machine...

I haven't tried it with the shareware, but to my knowledge, the
(gfx.wad?) only stores lumps for graphics like the console background,
status bar, etc. Qtest had no mip-mapped textures in the gfx.wad.

Further, QuakeEd will take a dump if the first lump in the wad that
you try to import is not the palette for the textures. So, keep that
in mind if you all try to generate your own wad2 texture files for

BTW, here is some info that I gathered last night with the id tools:

I re-built dm1.map into dm1.bsp using all three tools. These are the
results for my computer (and the command-line I used, just for reference):

qbsp: (qbsp dm1.map) : 130 seconds
light: (light dm1.bsp) : 63 seconds
vis: (vis dm1.bsp) : 367 seconds

This was on a P5-100 with 64M of ram running win95. I thought this was
pretty encouraging. BTW, for those who didn't know, dm1 is just test1
from qtest. Anyway, this isn't a huge level, but it isn't trivial either.
Unless I misunderstood, the above should have built full versions of all
repective parts. If you used the various short-cuts, or cut out some
of these all together, I think you bould get turn around in probably
2-3 minutes for most levels. If you work on the level in parts, as suggested
by Carmack I believe, it should be a lot less.

Anybody have any idea what's up with the QuakeC compiler? I would like
to start making some levels now ;)