Mark Jones wrote:
> One of the greatest disappointments to date of the move
> to more liberal (i.e. free software) licenses for systems
> like Hugs and GHC, is that it has done almost nothing to
> stimulate contributions to the implementations themselves
> from outside the immediate (and small) group of developers
> concerned. Compare this, for example, with the Linux
> community where the number of external contributors is
> often cited as one of the benefits of the development
> model used there. Of course, it may just be the size
> of our community, and the subject area: there's a much
> greater demand for operating systems than there is for
> lazy functional language implementations, and there are
> probably a lot more people with expertise in the former
> than there are in the latter.
I recently spent a few hours surfing, reading the fsf, gnu, linux
sites and found quite a few differences to haskell.org. I don't claim
that the lack of interest in contributing to FP software is solely due
to these but they surely reflect differences between the Linux/FP
communities (the FP community is not a Bazaar, more like a Cathedral
-- see Eric Raymond's classic):
1. If you look at www.gnu.org there is a clear list of ongoing,
unfinished, abandoned projects and a list of what people can (and
encouraged to) contribute. In other words you look at these pages and
pick your 'territory', two months later you post the results of you
efforts and others will start using it, take it apart, improve it
etc. [Aside: for a rather interesting discussion about the role of
'noosphere' see www.opensource.org/links.html].
On the contrary in FP land, there are announcements..."I added XXX to
{GHC, NHC, Hugs etc}, here it is, use it". Two months later some
documentation appears and people start complaining that extension is
not sound. There is no discussion what people are adding to FP
software, therefore there is no beneficial effect of 'peer review'.
Also, once something is in, some people will like it and it is harder
to remove it. Just consider the recent discussions regarding the
layout rule. I guess the roots of this aspect is in the way academia
works...
2. I don't know much about the gnu people, but Linus Torwalds made it
very clear in a recent interview that most changes to Linux are
expected in user space not in the kernel. He is basically saying that
the interfaces will not change fundamentally, you can develop your
contribution and (perhaps with minor tweaks) it will work. Contrast
this with changes to GHC from versions 2.10 to 4.04...there has never
been any guarantee that your piece of contribution will work in a
later release.
3. Despite all the claims that FP in general allows independent
software development, modularity etc the ongoing projects do not prove
this. You can't just add a magic pass in isolation...you have to
understand what ghc/compiler/prelude/TysWiredIn.lhs (or something like
that) is doing when all you wanted is to add deforestation. So the
general involvement/investment required to contribute is bigger than
necessary.
Just my 2 HK dollars...
All the best,
Laszlo
