|
Re: more of Roland's bad math (was Re: Movement Waits off?) (Community)
Re: more of Roland's bad math (was Re: Movement Waits off?) // Community
Jan 30, 2001, 6:49pm
I have a half-done build at 30504n 0w AW (stopped because of the gaps), I
may still have demonstration objects of this behind it. The gaps are quite
L A R G E and you can't miss it (if it's there
anymore)
--
Wing
This little spot is dedicated to my girl, Jessie.
WHAT TIME IS IT? GAME TIME!
Ravens are goin Festivus Maximus on Purple Sunday
AW Citizen 305004 "Wing"
bathgate at prodigy.net
eyemwing at teleport.com
ICQ #101207433
[View Quote]goober king <rar1 at acsu.buffalo.edu> wrote in message
news:3A76ED8A.6803EAA8 at acsu.buffalo.edu...
> Eep knows full well why he didn't get in the beta, and it has nothing to
> do with his (in)ability to "run circles around Roland".
>
> As for the gap problem, since it only occurs only out in the outer
> regions of AW, (in the 5-digit coord areas) you need to have a sizable
> world in order to see it. So you won't have to worry about it much in
> the smaller, private worlds. I myself haven't tested it, but then, I
> haven't built much in the outer regions either. With what little
> programming experience I have, my guess is Roland just didn't allocate
> enough significant digits for the coordinate system, so when you get
> into the higher numbers, they get rounded off a lot higher. (That, I'm
> assuming, is what he meant by "floating point rounding errors") But
> considering how much resources AW already uses up, would allocating a
> few more digits really help in terms of memory consumption? It adds up
> after a while :-/
>
> myrth wrote:
else
mainly
aw?
you
grounds
the
WRONG
everytime
the
hours,
it
>
> --
> Goober King
> Could help if they had made the whole thing in JAVA... ;P
> rar1 at acsu.buffalo.edu
|
|
