home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
The Arsenal Files 8
/
The Arsenal Files Collection #8 (Arsenal Computer) (1996).ISO
/
g_quake
/
vistime.zip
/
LEETVIS.TXT
< prev
next >
Wrap
Text File
|
1996-10-02
|
3KB
|
78 lines
LEET VIS
Current Version: Time Testing Version
-------------------------------------
standard disclaimer
-------------------
If this program causes your computer to have problems in any way, whether it
explodes or loses its cmos or whatever, its not my fault. By reading this
and using leetvis you understand and acknowledge this disclaimer.
Intro
-----
After working on this for a while I have finally got a pretty stable model.
I have examined and looked at the vis program for some time and tried to
understand its algorithms as thouroughly as possible. I found some
relationships between the data elements vis was processing and tried to
correlate them. Finally I did a curvefit on several key pieces of data and
got a pretty good model. In english this means that I have a vis time
predictor.
Note that this not for every possible case of vising. This model is just for
a full vis using level 2. Fast vises and other levels do not apply but the
prediction should give you a ballpark estimate. Also not that this may be
terrible for worse cases of vis. The vis time data is fairly straighforward
at low vis times but the worse cases can wildly fluctuating times and the
prediction of those cases is difficult at best.
Just remember that since the generation of data from vis itself depends highly
on the level geometry so does this predictor. As a rule of thumb for levels
with low critical ratios the vis time should lean towards the low side and
for those with higher (>2.5) critical ratios the times lean towards the high
side. Predicting your vis job time is quite an art as you can see.
Oh yeah for the issues of different types of systems and their vis times:
The program was calibrated with a 90 MHz P90 with 16 MB ram. Nothing fancy.
If you are on a 486/33 it will probably be a lot slower. Likewise on a
dual ppro200 the vis times will fall a bit lower :)
In general for cases of very small levels with few brushes the low estimate
is good. For cases of most of the levels I tried (telex, sphere, qchunks,
etc) the true times are a bit lower than the average. For levels with very
high critical ratios ( >3.1) the high estimate is good and many times the
true vis times are greater than the high.
current options for leetvis
---------------------------
leetvis prtfile.prt
it just takes a portal file and does its magic.
Thanks and greetz
-----------------
Thanks to GOD or else we wouldnt be here. Also thanks go out to iD
software because they love us almost as much as He does ;-) Also I would
like to thank Billy Zelsnack for his putting up with me and being such a
great help in my other projects. Special greets to Dave Taylor at crack.com,
Ryan Geithman at bossgames, and all my wonderful friends on irc.
Starving Coder Standard Beg for dollars section
-----------------------------------------------
LeetVis is FREEWARE. Use it and give it to your friends. If you like the
program and it gives you a wood, please feel free to make a $5 donation to
Feed The Children, the Red Cross, or the Salvation Army. There are those
who need it a lot more than you or I.
Complaints, Concerns, Comments, etc...
--------------------------------------
They all go to vile@dazed.nol.net or /dev/null depending on how angry you are
at me ;)
--------------------------------
.prts is .prts and pieces .prts.