On 18 Sep 1998, Will Partain wrote:
[cut]
> We (Glaswegians) tried to do a "Haskell library" in the
> early days, but not much came of it except for Stephen
> Bevan's numerical code. But that's no reason not to swing
> 'round and try again; after all, there should be *much* more
> Haskell code in existence than in 1991.
>
> Will
Please do! But I beg you to revise some of the library
from time to time, perhaps by contacting the original
authors? Surely they would happily make some improvements
-- granted some interest in their work. It is not enough
to say "we have it", it will be nice to assure that
it works and is well documented. Otherwise we have an illusion
of the library only.
I recently tried some of the code you mentioned (I do not
remember whether it was Stephen's or somebody else's),
but I am sorry to say that not all worked and some that did
was unfortunatelly too narrow for what I needed.
Case in question: LU decomposition of matrices.
A very basic task, but what I found in your repository
applies only to symmetrical matrices (LL decomposition).
Real life is not always so symmetric! :-)
Let me cite from the Eiffel practise (Again, but why
not if the practise is sound?).
No code is accepted to their repository unless it
passes certain tests pertaining to:
-- style
-- documentation
-- assertions (This does not apply to Haskell though)
-- quality
-- reliability (including assertions and error recovery)
A team that does it is (was?) not always the best suited
to the task but at least they are able to catch the
gross mistakes and give an author some advice how to correct
the problems. They are picky but not mean, and everything
is done in a friendly manner.
Who would not volunteer to be a subject of such a scrutiny
if this were to be done for the good cause?
Jan
