Need reasons to GPL Haskell implementation
The following message is a courtesy copy of an article
that has been posted to gnu.misc.discuss,comp.os.linux.advocacy as well.
Hi,
This is a long post, so I've divided it into sections---Preface,
Request for Feedback, Consideration, Conclusion, Appendix---to help
you figure out whether it's worth reading. :) TIA for any feedback and
I will summarize if I get enough info. Haters and lovers alike of the
GPL are invited to comment, criticize, flame, etc.
I. Preface
To those of you who do not know about Haskell, it is the most
important purely functional programming language. Several groups
worldwide are actively developing Haskell implementations. See
<http://haskell.org> for more info, such as on the implementations:
HUGS, GHC, Chalmers, NHC.
Again, to those unfamiliar with functional languages, consider the
place mathematics has had in advancing technology. Consider the
potential for good in a programming language with an exceptionally
strong mathematical basis, such as Haskell.
II. Request for Feedback
_Next week_ top developers (of HUGS and GHC) are meeting to consider,
among many other things, whether to move to the GPL. In the Appendix
is a license of one of the implementations, HUGS.
I have been asked to email them reasons why a Haskell implementation
should be GPLed, and I _ask you for feedback._
III. Considerations
(1) An important consequence of licensing is how amenable it is to
distribution, such as with the Linux distributions (RedHat, Debian,
etc.). Restrictive licenses such as HUGS (see part [b] of the
license) or unclear licenses (such as with the GHC implementation)
result in many fewer people not enjoying and using some important
software. I know that is why there is no Debian Haskell package yet.
(2) Naturally, one of the top concerns of the developers is that
Haskell is used as widely as possible. It seems that they do not care
whether that is via the sharing that the GPL encourages or via the
binary distributions that are so typical in the Windows community. In
my experience, I find that most users are oblivious to licenses,
though the Linux community seems to be more aware. Therefore, the GPL
would be quite suitable.
(3) In fact, although these implementations are available for several
platforms, there seems to be quite a bias towards Windows. For
example, there has been collaboration with Microsoft Research on
developing a Haskell animation system ("FRAN") that works (naturally)
only in Windows, and while the Windows graphics interface has been
actively maintained, there is none available for the X Window system.
Please understand, FRAN is great stuff, and I have nothing against
development for Windows---it's just a real shame that little is being
done for other OSes. At the very least a GPL might encourage more
development within the Linux community, for example. OTOH, developers
of proprietary software (such as Microsoft) would avoid software
licensed under GPL.
(4) Probably (3) is motivated partly because of the ubiquity of
Windows. But as a Linux user, I am deeply concerned when important
research occurs with proprietary, technically inferior OSes, while
free, excellent ones such as Linux and *BSD take a back seat. Please,
no OS war; I just have yet to see that Windows [95 and NT] works as
well or provides as much freedom or flexibility.
(5) Regarding advanced languages in general, GUILE seems to be ahead
as far as technical merit and licensing go, but Haskell certainly has
some important advantages over Scheme. Improving the license of a
Haskell implementation would make it a favorable alternative to
GUILE.
(6) Is there any license other than the GPL that better ensures that
important software technology remains freely available?
IV. Conclusion
Some believe the GPL and functional languages lead to the development
of good, reliable software.
I feel that a GPL on this important software will (A) encourage the
use and development of Haskell, and (B) help free OSes maintain a
place in an important area of CS research.
Now, (B) is all well and good and encourages us to care but of course
is irrelevant to Haskell implementors. (A) is what I need help on.
Can anyone help?
V. Appendix
______________________________________________________________________________
___ ___ ___ ___ __________ __________
/ / / / / / / / / _______/ / _______/ Hugs 1.4
/ /___/ / / / / / / / _____ / /______
/ ____ / / / / / / / /_ / /______ / The Nottingham and Yale
/ / / / / /___/ / / /___/ / _______/ / Haskell User's System
/__/ /__/ /_________/ /_________/ /_________/
Copyright (c) The University of Nottingham and Yale University, 1994-1997.
Bug reports: hugs-bugs@haskell.org. Web: http://www.haskell.org/hugs.
CONDITIONS OF USE, DUPLICATION AND DISTRIBUTION (*)
Permission to use, copy, modify, and distribute Hugs for any personal
or educational use without fee is hereby granted, provided that:
a) This copyright notice is retained in both source code and
supporting documentation.
b) Modified versions of this software are redistributed only if
accompanied by a complete history (date, author, description) of
modifications made; the intention here is to give appropriate
credit to those involved, while simultaneously ensuring that any
recipient can determine the origin of the software.
c) The same conditions are also applied to any software system
derived either in full or in part from Hugs.
No part of Hugs may be distributed as a part or accompaniment of any
commercial package or product without the explicit written permission
of the author and copyright holder. The distribution of commercial
products which require or make use of Hugs will normally be permitted
if the Hugs distribution is supplied separately to and offered at cost
price to the purchaser of the commercial product.
In specifying these conditions, our intention is to permit widespread
use of Hugs while, at the same time, protecting the interests, rights
and efforts of all those involved. Please contact the author and
copyright holder to arrange alternative terms and conditions if your
intended use of Hugs is not permitted by the terms and conditions in
this notice.
While Hugs has much in common with Gofer (from which it was originally
derived), there are also some significant differences between the two
systems. For example, Hugs conforms closely to the Haskell standard
while Gofer was intended as a more experimental system. As a result,
any and all rights previously conferred for the use, duplication, and
distribution of Gofer do NOT automatically carry over to Hugs.
NOTICE: Hugs is provided "as is" without express or implied warranty.
(*) For the purposes of this document, the word "Hugs" refers both to
the software and its accompanying documentation.
______________________________________________________________________________
--
TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to
debian-user-request@lists.debian.org .
Trouble? e-mail to templin@bucknell.edu .
Reply to: