From: Reid Sweatman (borderland_at_[hidden])
Date: 2000-12-21 17:05:40
I'd be interested. I've got my own quaternion class, but I'm always willing
to see if someone can show me something better, and it'd be nice to have it
in Boost. Like Gary, I use them a lot in games (ex-Sierra myself). It
would be nice if it could build on stuff already in the STL or Boost. I'd
really like to see a general high-quality *fast* numeric library in Boost.
As I said, I've got my own stuff, but it isn't exactly portable, being an
odd mix of generative programming techniques that use valarray, other
containers, and custom assembler that's CPU-specific. Else I'd have offered
it to Boost by now.
I can just guess, though, that someone will shortly tell me that I should
just use Blitz++ or something <g>. Well, I'm keeping my eye on it, but my
currently favorite compiler (no, not that one, so don't start <g>) won't
compile parts of it. But I well understand Dave's comment, because
*accurate* / *fast* numeric stuff takes an expert to get right.
Just out of curiosity, what are "octonions?" (I can sort of guess, but I'm a
mathematician by training, and I ain'ta never heerd of 'em). Could you
suggest a book or even better, an online paper that gives the gist? Thanks.
> -----Original Message-----
> From: Hubert.Holin_at_[hidden] [mailto:Hubert.Holin_at_[hidden]]
> Sent: Tuesday, December 19, 2000 5:12 PM
> To: boost_at_[hidden]
> Subject: [boost] Interest for quaternions (and octonions)?
> Paris (U.E.), le 20/12/2000
> Some time ago I posted an implementation for quaternions on
> comp.std.c++, and was told boost was a more appropriate venue for this
> kind of submission.
> I would therefore know if there is interest in that tool
> among the boost constituents.
> I have read the parts of the boost website relative to
> library submission, and guess massaging my file for compliance with the
> guidelines would only entail minor (cosmetic) work. Copyright is not a
> problem, but I currently can only test on the Mac version of
> I should mention I also have an implementation of octonions
> in the work, and would also like to know if there is also interest in
> it (I guess it must be quite minor with respect to that possibly
> generated by quaternions ;-) ).
> Hubert Holin
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk