From: Peter Schmitteckert (boost) (boost_at_[hidden])
Date: 2004-07-06 13:58:09
On Tuesday 06 July 2004 17:31, Jeremy Siek wrote:
> Yes... the term "matrix" has a multitude of meanings, all of which
> are confusingly similarly but slightly different. However, the current
> of conversation is a linear algebra library. Thus it makes sense for
> the purpose
> of this conversation to restrict the meaning to that of traditional
> linear algebra.
This was not clear for me from the beginning, since I came aware of the discussion
due to cross-posting to ublas-dev.
> Believe me, I'm not trying to exclude physicists or anyone. I'm highly
don't worry, I think nobody ever assumed this,
> favor of there being lots of numerical libraries in Boost. There can
> be many different concepts with the name "Matrix", they'll just be in
But we should avoid an explosion of matrix libraries, since users than
have to evaluate them, to find the one fit their needs best.
> different namespaces. All I'm saying is that the concept named "Matrix"
> in the linear algebra namespace should have the standard linear algebra
> meaning and syntax.
However, it would be nice to be aware of usage beyond 'standard linear algebra',
and whenever possible, one should try to enable those applications.
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk