Quake3World

http://www.eurekalert.org/pub_releases/ ... 091305.php

hahaha take that morons...

I like the fact I haven't yet seen the actual formulas :\

oh, and stop posting freakaloin

Fascinating. What's great is that Goof clearly has no idea what that article means.

it means u suck cock..i do know that...

Oh, something else you don't know then.

Freakaloin wrote:it means u suck cock..i do know that...

good guess, but way off the mark.

if this is true it would lead to improved 3d maths performance fps games surely

thx...

is this info available outside of that book?

no...its classified...

4days wrote:is this info available outside of that book?

i haven't been able to find it anywhere, since i'm actually quite interested in this. seems like he's just being greedy with it all. it's becoming easier and easier to spot the assholes who are slowing down the evolution of humanity

you can pick up the first chapter at
http://web.maths.unsw.edu.au/~norman/book.htm

having a look at it now (with the ejit's guide to algebra in another window). the missus is a bit of a mathematical genius, but if i can avoid asking her to explain it to me then i will

4days wrote:is this info available outside of that book?

http://science.slashdot.org/comments.pl ... d=13584486
http://science.slashdot.org/comments.pl ... d=13584500

good posts on it.


Basically, it's trig but without the sine() etc. functions. Nothing revolutionary, just different.

jesus...seems Freakaloin is the new king of spam...ouch you go girl

neh wrote:if this is true it would lead to improved 3d maths performance fps games surely

Its fundamentally incomplete as compared to trig. It eliminates any notion of "between-ness" which is quite important for many applications and most 3D shooters.

hax103 wrote:

neh wrote:if this is true it would lead to improved 3d maths performance fps games surely

Its fundamentally incomplete as compared to trig. It eliminates any notion of "between-ness" which is quite important for many applications and most 3D shooters.

oh well

i didn't click on the link because i already know i won't understand any of it

maths

AWESOME NEWS!!!! THIS IS FUCKING AMAZING!!!

(seriously)

For people who think he's sitting on the information in order to get you to shell out 78 AUD for his book, here's Chapter 1 (which he provides freely)

http://wildegg.com/papers/Chapter1.pdf

hax103 wrote:its fundamentally incomplete as compared to trig.

In what way?

hax103 wrote:Its fundamentally incomplete as compared to trig. It eliminates any notion of "between-ness" which is quite important for many applications and most 3D shooters.

Sure? What kind of "between-ness" are you talking about?

Wow, this is a pretty amazing new approach. It seems that the relationships to classical numbers and equations thoroughly back up this guy's research.

If you haven't read any of it you can see the Preface, Intro, ToC, and 1st Chapter here:
http://web.maths.unsw.edu.au/~norman/book.htm

:icon14:

Pext wrote:(seriously)

programmers of 3d engines have been doing this for years.

yeh... but from a mathematicians point of view this is very interresting, as trigonometric geometry is used in a multitude of fields, not only restricted to pure geometry.

concerning between-ness:

i think the problem you are talking about is wether, from a given point of view, something is in line of sight - so we can abstract this probles as:

let S, A, B, Q be 2d-points - is point Q visible between A and B from point S?

in classical geometry, this is simple: you simply test if the absulute angle of line SQ is between the absolute angles of SA and SB.
But you have to sort the angles of SA and SB and, accordingly, do the 'between?' test

in rational goemetry it's a bit different. You check:
s(SA, SB)>s(SA, SQ) ?
and
s(SA, SB)>s(SB, SQ) ?

In this case, the sorting of absulute angles is not needed anymore. On the other hand, you get the problem, that for every point valid point Q1 the point Q1', that is mirrored at S, is also valid.

You either could do sign checks for x_Q-x_S ?= x_A-x_S , etc - but this is a bit of a shitty solution.
Or maybe there is something to be done with s(SA,SB)>s(QA, QB)...