home *** CD-ROM | disk | FTP | other *** search
- Hi,
-
- AAAAAAAAARHHHHHHHHHHHHHHHHHHH.............
-
- Phu.... Okey. I have just received a demo of the new texture
- mapped version from Doug but I'm having some troubles with it.
- :-(((((
-
- This time NeST have been kind to not mess with silly LF corruptions,
- this time it have done something realy uggly, or it's something wrong
- with my ZIP program using Unix. The uudecoded file looks very good
- and it buildes into a nice ZIP file, but then unzip complains about
- missing bytes (90 of them). It does unzip it, but I'm not sure if
- it manage to do it correctly. I will send both the original uue files
- and a new archive that will hopfully work correctly.
-
- This archive does not include any source, just executables and is far
- from ready. Anyways it's faster in 320x200 then DEE was in 80x50. :-)))
-
- Here is the text from the readme file, if someone have got troubles
- with it you can atleast read this. ;-) I will ask Doug if he can send
- a new one through Titan Design insteed (they have proper internet access).
-
- //Magnus Kollberg
-
- ------------------------------------------------------------------------
-
- Doom Environment Simulator v1.7alpha
- ------------------------------------
-
- Well, here it is. It's far from finished, but it works - and gives a good
- idea of how fast the final engine will be after all things are considered.
-
- There is quite a bit of work to be done on a texture cache before we can
- use multiple textures - anything short of this is a waste of programming
- effort. Virtual memory would work, but it's kinda overkill and you can't
- hope to influence it's efficiency if you have no tangible cacheing system
- to work on.
-
- Since I am completely unable to track down any useful information on the
- mathematical formulae employed inside Doom & it's clones for the perspected
- texturemapping, I was forced to develop my own formulae which are likely
- to be nothing like the real thing in shape or form. The point is that the
- result is the same. If anyone can enlighten me on the Doom perspective
- mapping problem, please do. Until then (and perhaps even afterwards) I'll
- stick to my own home-grown methods.
-
- The complexity of the DSP module has now broken broken all previous
- boundaries in nastiness since the heapload of maths has been added for the
- texturemapping of floors & walls. As a result, the 68030 sources have become
- fragged and need some work before anyone is likely to be able to get anywhere
- with it. I have given up hope on making the DSP code readable. It's beyond
- saving, and I doubt it's worth the massive work required to comment it and
- leave it open for modification. From now on, the DSP proccessing unit of this
- project becomes a 'black box'. Some may find this irritating, but I warn
- those same people that 60k of uncommented, straight DSP logic & math formulae
- is very BAD NEWS.
-
- Important modifications since v1.32alpha:
-
- > The NodeInCone function is now on the DSP.
- > A new DSP function has been employed to defrag the floor runs.
- > Rotational, perspected mapping function for floors & ceilings.
- > Distance shading for floors & ceilings.
- > Perspected mapping function for walls.
-
- The code badly needs optimised, and there is a small about of work to be
- done before the walls can be distance-shaded, so the speed is likely to
- improve later on.
-
- keys:
-
- 1 & 2 rise & drop
- 3 toggle high / low detail (RGB monitors only so far)
- TAB map on / off
- Arrow keys turn / forwards / back
- ALT strafe / sidestep mode
- - decrease window
- + increase window
-
- Anyway, don't read TOO much into the framerate yet, as there is a bit of
- work left which will slow things down, and a lot of optimisation which will
- have the opposite effect. We will just have to wait and see. This should give
- everyone a good estimate, but not a hell of a lot more than that.
-
- Have fun....
-
- Doug @ BSS <dlittle@nest.demon.co.uk>
-
-