From: Brinkerhoff <brinks AT ainop DOT com>
Newsgroups: comp.os.msdos.djgpp
Subject: 3D clipping in Allegro (sorta)
Date: Fri, 06 Mar 1998 22:06:32 -0800
Organization: Dave's Machine Service
Lines: 50
Message-ID: <3500E3E8.54BA@ainop.com>
NNTP-Posting-Host: mast65.ainop.com
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
To: djgpp AT delorie DOT com
DJ-Gateway: from newsgroup comp.os.msdos.djgpp
Precedence: bulk

This idea may have been done before, but I've never heard of it being done 
before.

I haven't tried it, but it seems like it should work, and it seems like it is 
mathematically correct.

I came up with the idea to make walls for a 3d engine, but it should work 
with any polygons.

The idea is that with Allegro, polygons that have points that all have a 
positive z value will all be drawn properly (pretty much).  I have searched 
the net for info on clipping, and frustrum (is this the correct word?) 
clipping is a little too complicated for me, and it seems to me that it 
is too much work.  I haven't actually sat down and compared or benchmarked 
it, but my idea seems that it should be faster. 

Suppose you have a wall (a four sided poly).  Lemme try a diagram to help 
explain.

                  (d,e,f)
              | / 
              |/|
              | / (j,k,l)
             /|/
     (a,b,c)/ |
            |/|
            / |
     (g,h,i)
              ^(xy plane, where z=0)

Notice that one side of the poly is in the negative z region.  So, find the 
slope of the lines (a,b,c)-(d,e,f) and (g,h,i)-(j,k,l).  Using these slopes, 
find where the lines intersect the xy plane (by finding an x slope and a y 
slope, if this makes sense to you), giving you points (a',b',0) and 
(g',h',0).  When you draw your poly, use these points, instead, making all 
your points end up in front of the xy plane instead of behind it (btw, what i 
call the xy plane is the plane that contains both the x and y axis where 
z=o).

Since I do not understand completely how allegro texture maps, this is where 
the difficulty begins.  but I was thinking that if you used the distance 
formula to find the distance between (a,b,c) and (a',b',0), you could just 
take this value and add it to your u value for texture mapping.

Im open for criticism.  If anyone can find a way to improve upon this without 
destroying the original idea or making it too complex, id like to know.  id 
also like to know if my ideas for texture mapping would work, and if they 
wont, what will?

Glen Ellis Brinkerhoff