Re: Math library

[Darren J Wilkinson]

Keith Hopper wrote:
> Seriously, count me out as far as the algorithms and design work go.  What
> I suspect might be useful is a look at not re-inventing the NAG library!
> I am quite happy to review written code - even if someone gives me test
> programs to do some testing.

I guess we are all united in despising the NAG library! Has anyone else here
looked at the GNU Scientific Library (GSL) (http://sources.redhat.com/gsl/)?
I've been using it quite a bit for scientific computing using C. I generally
quite like it, and it has been designed so as to be relatively easy to link to
from a VHLL such as Sather. So one possibility for a the future of the Math
library is to make the interface consistent with the GSL, and provide users
with an option of using a pure Sather implementation (which has certain
attractions), or the GSL itself, which is probably less buggy and much
faster/more efficient. Does anyone have any views on this? Would people rather
just stick with pure sather implementations wherever possible, or are people
happy with simply providing nice clean OO sather interfaces to an existing C
library? There are clearly some big decisions to be made at some point, so I
guess posting views here is as good a way as any of getting some input.

However, in the short term I just want to get the Math library for 1.3 working
as well(?!) as the 1.2.1 library. I've already made some progress in this
direction, and hope that my fixes will make it into the next release.

>      What on earth (or anywhere else) is CVS????? I know about CSV but I
> don't think you mean that.

I'm not a CVS expert myself, but I think it stands for "Concurrent Versioning
System" (or something similar). It is basically just a software database which
can be made accessible over the internet, which allows several developers to
work on a software project simultaneously.

-- 
Darren Wilkinson  Email:  address@hidden
                    WWW:  http://www.btinternet.com/~darrenjwilkinson/