[Haskell-cafe] Darcs users [was: New book: Real-World Haskell!]

[Laurent Deniau]

Jon Harrop wrote:
> On Wednesday 30 May 2007 06:58:36 Ketil Malde wrote:
>> On Tue, 2007-05-29 at 14:05 -0500, Doug Kirk wrote:
>>> I *want* people (and companies) to move to Haskell
> 
> As a complete noob considering making a commercial venture into
> Haskell, may I ask what people's opinions are on this? Are there many
> Haskell products?
> 
> Our expertise is in scientific, numerical and graphical computing.
> Our product catalog should elucidate this:
> 
> http://www.ffconsultancy.com/products/
> 
> I'd like to know how feasible it would be to rewrite some of these
> products in Haskell. For example, our time-frequency add-on for
> Mathematica might sell a lot better if it were a standalone
> cross-platform GUI application or even a web application. Is it
> feasible to write such a thing in Haskell? Are any core Haskell
> libraries non-free for commercial use?
> 
> Also, would anyone be interested in a Haskell for Scientists book,
> following our OCaml for Scientists and forthcoming F# for Scientists
> books?

I didn't read these books but I would if one exists on Haskell. But
I expect that you will first have to convince scientists that Haskell 
can be efficient in numerical analysis (both memory and speed). My guess 
is that it can be efficient if missing efficient data structures are 
provided. This has been solved for strings so I guess it can be done for 
other data structures as well but it seems to be a huge work. If the 
target is only to wrap the GSL or equivalent, then I expect that it will 
only be considered as a curiosity.

a+, ld.