caml-list - the Caml user's mailing list
 help / color / mirror / Atom feed
From: Sven LUTHER <luther@dpt-info.u-strasbg.fr>
To: Brian Rogoff <bpr@best.com>
Cc: Jacques Garrigue <garrigue@kurims.kyoto-u.ac.jp>, caml-list@inria.fr
Subject: Re: [Caml-list] CDK license
Date: Wed, 6 Jun 2001 09:40:57 +0200	[thread overview]
Message-ID: <20010606094057.B4764@lambda.u-strasbg.fr> (raw)
In-Reply-To: <Pine.BSF.4.21.0105301958510.10735-100000@shell5.ba.best.com>

On Wed, May 30, 2001 at 08:11:56PM -0700, Brian Rogoff wrote:
> On Thu, 31 May 2001, Jacques Garrigue wrote:
> > On the same line of thought, the ocaml compiler is released under the
> > QPL, which is not compatible with the GPL.
> > This means that you cannot build a toplevel including any library
> > under the GPL, since it would be in contradiction with either of the
> > two licenses.
> > At the very least, it seems necessary to add a clause to the GPL,
> > saying that linking to QPLed libraries is allowed, just as RMS himself
> > suggested for KDE software.
> 
> At this point, I'd suggest that we _really_ need to consult a lawyer who
> is familiar with intellectual property law and the GPL. 

First, i am no lawyer, but as debian developper, and packager of ocaml and
some other ocaml stuff for debian, i have been exposed to this kind of stuff a
lot, so here is my opinion on it.

> As far as libraries go, I think the LGPL is a fair compromise between the 
> really dedicated RMS followers (once affectionately referred to as
> "license ayatollahs" on this very list :) and those who are willing to
> tolerate a variety of kinds of software, including proprietary. I
> understand the reasons for going GPL instead of LGPL ("resistance is
> futile, prepare to be assimilated, or don't use this code") but if it's 
> going to be that way then I don't want my Consortium dues to fund work
> on the CDK. 

Yes, the LGPL is a good choice, but still it has some problems with regard to
ocaml programs/libraries.

I think the important part is article 5. of the LGPL :

------------------------------------------------------------------
  5. A program that contains no derivative of any portion of the
Library, but is designed to work with the Library by being compiled or
linked with it, is called a "work that uses the Library".  Such a
work, in isolation, is not a derivative work of the Library, and
therefore falls outside the scope of this License.

  However, linking a "work that uses the Library" with the Library
creates an executable that is a derivative of the Library (because it
contains portions of the Library), rather than a "work that uses the
library".  The executable is therefore covered by this License.
Section 6 states terms for distribution of such executables.

  When a "work that uses the Library" uses material from a header file
that is part of the Library, the object code for the work may be a
derivative work of the Library even though the source code is not.
Whether this is true is especially significant if the work can be
linked without the Library, or if the work is itself a library.  The
threshold for this to be true is not precisely defined by law.

  If such an object file uses only numerical parameters, data
structure layouts and accessors, and small macros and small inline
functions (ten lines or less in length), then the use of the object
file is unrestricted, regardless of whether it is legally a derivative
work.  (Executables containing this object code plus portions of the
Library will still fall under Section 6.)

  Otherwise, if the work is a derivative of the Library, you may
distribute the object code for the work under the terms of Section 6.
Any executables containing that work also fall under Section 6,
whether or not they are linked directly with the Library itself.
------------------------------------------------------------------

What all this is about is that you can do dynamic linking with the library and
not release your program source code.

That said, ocaml is not a dynamic linking language yet, and as such i don't
know up to what point the :

  only numerical parameters, data structure layouts and accessors, and small
  macros and small inline functions (ten lines or less in length).

Could be related to the way ocaml uses modules and do linking. Could someone
more familiar with the ocaml binaries internal give more light here ?

Now, there are 2 possibilities :

  1) We consider that the ocaml modules and linking stuff is considered such as
     Article 5 requires. Then there is no problem, and we can use the LGPL
     without major problem. But it would be nicer, when giving a LGPL license
     to your code, to make it clear that we consider it as such, saying for
     example :

     "This code is covered by the LGPL. Notice that we consider the ocaml
     linking and modules hadnling as consisting of only numerical ..."

   2) Ocaml linking is more than just the above. Then you can still solve it
      by :

     2.1) state in the license that you make an exception for ocaml linking,
     something like :

     "This code is covered by the LGPL with the additional permision that you
     may use the object file in an unrestricted manner, when linking to other
     ocaml objects."
     
     2.2) comply with article 6 a), which says that you must also provide a
     complete machine readeable version of your program (the .cm* files) so
     that the user can modify the library and relink the program with the
     modified library. Now you don't need to distribute the files, you can
     resort to make them available (for at least 3 years) in either a
     downloadeable place, or in sending it for a cgharge no more than the cost
     of distirbution to anyone who request it. 


And naturaly, don't forget that you can issue any code with any number of
licences that you want, provided _all_ the authors agree on it. If you plan to
do so, best is to check with the consent of anyone who makes contributions to
your work, before integrating the patches. Or keep a dual source tree, as was
done with mozilla/netscape and some others.

Ok, that is most of what i have read from the LGPL stuff, and which applies to
ocaml. 

> > Another remark is that lablgtk-1.2.0 contains a COPYING file, saying
> > that the library itself is LGPL, examples are more or less public
> > domain, and applications are _not_ open source. Claiming that all this
> > is GPL is clearly wrong. This COPYING is not there, and no README file
> > either, which is the only documentation for lablgtk :-)
> > 
> > (This is not a rant: I am the one who didn't check)

erm, ...

1) about the public domain examples, best would be to add a COPYING file there
stating it, because public domain is less restrictive than the LGPL.

2) about the application not being open source, what are they. I cannot
possibly continue to include them in the debian package if their license
situation is not clarified. Will they come under a closed source licence ? If
yes, why, does it make any sense to do so ?

3) IANAL, but it seems to me that since you shipped lablgtk claiming that all
of it is LGPLed, at least the versions that where such released are released
under the LGPL, you cannot go back. You can change the licence though, but i
guess anyone could continue work on the LGPLed version of it. I could be
speaking nonsense though, i did not check the exact wording of it.

Friendly,

Sven Luther
-------------------
To unsubscribe, mail caml-list-request@inria.fr.  Archives: http://caml.inria.fr


  parent reply	other threads:[~2001-06-06  7:38 UTC|newest]

Thread overview: 25+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2001-05-30 19:06 Brian Rogoff
2001-05-31  1:05 ` rbw3
2001-06-06  7:05   ` Sven LUTHER
2001-06-06  7:42     ` Sven LUTHER
2001-05-31  2:27 ` Jacques Garrigue
2001-05-31  3:11   ` Brian Rogoff
2001-05-31  7:46     ` Fabrice Le Fessant
2001-06-06  7:40     ` Sven LUTHER [this message]
2001-06-06  8:36       ` reig
2001-06-06  8:51         ` Sven LUTHER
2001-06-06  9:01           ` [Caml-list] question about modules Mark Wotton
2001-06-06 10:01             ` Markus Mottl
2001-06-06 13:17               ` Mark Wotton
2001-05-31 22:05 ` [Caml-list] CDK license John Max Skaller
2001-06-06  9:44 Dave Berry
2001-06-06 10:04 ` Sven LUTHER
2001-06-06 15:44 ` Stefan Monnier
2001-06-07  1:25   ` Jacques Garrigue
2001-06-08 12:59     ` Sven LUTHER
2001-06-07  9:03   ` Joerg Czeranski
2001-06-06 10:24 Dave Berry
2001-06-06 16:24 ` Brian Rogoff
2001-06-08 13:27   ` Sven LUTHER
2001-06-08 15:35     ` Brian Rogoff
2001-06-08 13:24 ` Sven LUTHER

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20010606094057.B4764@lambda.u-strasbg.fr \
    --to=luther@dpt-info.u-strasbg.fr \
    --cc=bpr@best.com \
    --cc=caml-list@inria.fr \
    --cc=garrigue@kurims.kyoto-u.ac.jp \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).