Re: [boost] [gsoc] Interest check for 3d geometry proposal

Subject: Re: [boost] [gsoc] Interest check for 3d geometry proposal
From: Emil Dotchevski (emildotchevski_at_[hidden])
Date: 2009-03-27 12:26:27

On Fri, Mar 27, 2009 at 8:55 AM, Michael Fawcett
<michael.fawcett_at_[hidden]> wrote:
> On Fri, Mar 27, 2009 at 11:50 AM, Phil Endecott
> <spam_from_boost_dev_at_[hidden]> wrote:
>> Kornel Kisielewicz wrote:
>>
>>> Also, in the case of 4d vectors, you'd have 5 values, which
>>> doesn't scale well ( 4*32 bits = 128bits which is nice -- an addition
>>> would ruin this niceness ).
>>
>> Multiplication by 4 faster than multiplication by 5?  Maybe, but probably
>> premature optimisation :-)
>
> I think he meant that an array of four 4d vectors would fit nicely in
> the 128 bit registers of a say, a Playstation, whereas an array of
> four 4d vectors with an additional size member would not.

Comments like "oh, a redundant size in the vector won't hurt, would
it" indicate that this isn't a game development mailing list. :)

I must say however that I'm not very optimistic about writing a Boost
library for game engine use. What's the next step, a Boost game
engine, maybe? :)

The main issue is that most software developers don't need such a
library; only game developers do, but you can't get them to agree on
the details. I have worked in the game industry long enough to see
that not only each team has their own "math" library that's slightly
different than everyone else's, but also many teams periodically end
up rewriting it. I've posted some more about this topic on my blog:
http://www.revergestudios.com/reblog/index.php?n=ReCode.MathLibrary.

Emil Dotchevski
Reverge Studios, Inc.
http://www.revergestudios.com/reblog/index.php?n=ReCode