Re: SD: RE: New operating system...


[Prev][Next][Index][Thread]

Re: SD: RE: New operating system...





>The big question here is, why? I haven't needed 10k of RAM in any
>assembly program, and I don't forsee any program needing 10000 numbers
>stored (or 1000 floating pt. nums).  

Sure this is only a natural question....  Quit simply I have had several 
rather ambitious projects in mind, but have always been discouraged from 
making them, because of having to abide by the TI-OS's rules.  And the 
proof that i am not the only one to feel this way is the simple fact 
that no one has come up with any programs that use more than the ~10K 
allotment.  It is an artistic limitiation to say the least.  Here is one 
project that I want to do, but becasue of the current OS, I find it hard 
to imagine how to implement correctly:  A 3-bit 3d rindering program 
with a full Zbuffer.  With he help of a friend of mine, I have come up 
with a recursive triangle drawing routine that will draw any sized 
triangle, by building it form a data bank of pre-drawn triangles.  This 
data bank itself takes up 4Kbytes. The good thing about this routnie is 
that in terms of speed, it favors small triangles.  Small triangles 
require no calculations to draw.  And slightly larger triangles only 
experience a small recursive load.  So objects with lots of polygons 
would actually be favored by this routine over 3d objects with few big 
polygons.  Then there are the Cos/Sin tables, and the Multiplication 
tables.  Plus there needs to be room somewhere to actually load the 3d 
Data so that "work" can be done on it such as rotational transformations 
and so forth.  Plus I might find it helpfull to keep a bank of polygon 
normals.  3-bit grayscale, and a 6-bit z80 with 128kbytes of ram should 
easily be able to handle such a program, but with the built in TI-OS in 
the way, it is almost impossible to create such a proram that can 
co-exist with the TI-OS.  Now I'm sure that there is probibly a way to 
make this proarm so that it does co-exist, but personally that is too 
much trouble to try and do, when it would be much easier to write the 
prgram for a better platform.  And basically a platform of this type 
would open the door to more extensive graphical applications in general.  
Now I am not familure with the concept behind raycasting so I do not 
know if more memory would help that out or not.

     Another program I wanted to write was a SCHEME(LISP like language) 
interpretor.  I feel confident that the ti-86 can handle such a program, 
but with such harsh limits on memory usage, I jsut could not find room 
to load all the built in functionality that would make the language 
useable.  A platform that would allow for code segments to go anywhere 
would make a SCHEME interpreter much more phesable.

     As for the incompatibility issue..  Sure.. I grant that this would 
be a pitfall.  But personally I never had a calculator in highschool.  
Of course I may have gone to one of the less intensive highschools I 
don'tknow, but I never really found a graphing calculator necessary.  
The only reason I got one was for the Assembly stuff anyway.  Maybe I 
should ahve gone with another device if that is all I wanted, but at the 
time, I did not know of any otehr devices in a resonable price range.

Later,

David E. West

______________________________________________________
Get Your Private, Free Email at http://www.hotmail.com