[clfswm-devel] CLFSWM license

Sat Jan 7 00:41:17 UTC 2012

madnificent at gmail.com writes:

> Oh, just stating:
>
> I'm fully open towards a mixed license thing.  IE: in the WebOS
> community, all patches were published under the MIT license (even
> those which extended the linux kernel), so as to allow for HP to take
> over these patches if it was good for them.  That basically yielded a
> mixed licensing strategy.  All code we wrote was MIT and could thus be
> incorporated without a legal team, though some of the stuff which
> happened on the linux kernel ended up being upstreamed as GPL, because
> the Linux kernel repository is GPL.  In the same way, I wouldn't mind
> it if CLFSWM would be GPL for all pieces of code which were taking
> from other GPL projects (if anyone would care enough to open CLFSWM up
> to MIT, they'd have to rewrite those portions into something else),
> and MIT for the specific portions.  That would at least yield a system
> which can be opened up in the future if sufficient interest would be
> there to do so.
>
Oh, here I clearly prefer one licence (GPL, BSD or another free one)
than a mix which complicate the contributions understanding.

best regards,

Philippe

>
> Best regards,
>
> Aad Versteden
>
> On Sat, Jan 7, 2012 at 12:44 AM, madnificent <madnificent at gmail.com> wrote:
>> Hello Philippe Brochard,
>>
>>
>> My main issue with this is that all code which is ran in the same lisp
>> image will need to be placed under the GPL.  That being said: placing
>> things in the contrib directory doesn't solve anything.  It might
>> solve something in a C world where things are linked differently, but
>> the trick doesn't work in Lisp.  If not for that, the GPL would still
>> not be to my liking, but it would certainly be acceptable for this
>> sort of software (and probably not only this).  The LLGPL tries to
>> solve this (though I must say that it hasn't had to stand up in court
>> either and it still scares me to death as I probably still
>> misinterpret it).
>>
>> For what it's worth: My stance against the GPL is not against free
>> software, I vastly prefer free software over proprietary software.  As
>> such I use free/libre Common Lisp implementations (my preference goes
>> out to SBCL and CCL).  I do *not* consider the GPL to yield free
>> software.  The legalese is not at all what it means for computer
>> scientists, if it can mean anything to us.  I kindly suggest you take
>> a look at [1], a section of the mails between rms and the CLISP author
>> in order to understand why the GPL doesn't work together with Common
>> Lisp.  The main issue with that in my mind wasn't even that I found it
>> to be unfair, it was that judges read a different license when they
>> see the same words.  The GPL can be seen as stating "the code must be
>> free".  The MIT/BSD license can be seen as stating "the programmer
>> must be free".  I prefer to be free myself, instead of having code
>> which is forced to be free.  Furthermore the moment at which code
>> needs to be shared is complete madness, it allows for easy misuse
>> (look at Google and TiVo for instance).  Therefore, in my mind,
>> there's little good to the GPL aside from the marketing mechanism
>> behind it.  Again, I do like sharing code.  I often do!  I like the
>> mindset which says that, together, we can get further, faster.
>> There's one world and we programmers should stick together for most of
>> it.  However, having a long and complicated license is likely not the
>> way to go, small companies can't afford to analyse or defend it.  If
>> there were a way to go, it certainly wouldn't be GPLv2.
>>
>>
>> Best regards,
>>
>> Aad Versteden
>>
>>
>> [1] http://clisp.cvs.sourceforge.net/viewvc/clisp/clisp/doc/Why-CLISP-is-under-GPL
>>
>>
>>
>> On Sat, Jan 7, 2012 at 12:02 AM, Philippe Brochard
>> <pbrochard at common-lisp.net> wrote:
>>> madnificent at gmail.com writes:
>>>
>>>> Hello,
>>>>
>>> Hi.
>>>
>>>>
>>>> I was looking at CLFSWM and I find the concept interesting.  I've
>>>> hacked the source code so the symbol manipulation works correctly in
>>>> most cases (allegro's modern mode will still crash it when changing to
>>>> another WM though).  I have more hacks planned.  However, I'm not at
>>>> all comfortable with the GPL license in a lisp system.  As shown with
>>>> CLISP, it is a recipe for legal issues.
>>>>
>>> First, thanks a lot for your interest in CLFSWM.
>>>
>>> Any hack and bug fix is welcome!
>>>
>>>
>>>> I personally release all code under the BSD license as I, as a
>>>> programmer, understand the license and know what I'll be allowed to
>>>> do.  I also value the fact that the code which I have written can be
>>>> used in the companies in which I work.  I know I will not get sued for
>>>> odd things I don't understand (again, read the mailing between RMS and
>>>> the CLISP author to see how lawyers's reasoning is different from
>>>> ours).  If you really want an obligatory (and to my view
>>>> condescending) license, I'd be willing to contribute under the LLGPL
>>>> (the Lisp Lesser Gnu Public License) which is most likely what you
>>>> intend to have if you are convinced the GPL is the only way to go.
>>>> For what it's worth, in any company I've  worked in so far, using and
>>>> contributing to BSD code was praised, GPL not so much (due to legal
>>>> issues).
>>>>
>>>> Though it's true that CLFSWM probably won't survive in a business
>>>> context, it makes all the more sense to publish it in a liberal
>>>> license which will not give either of the devs problems if something
>>>> would happen.
>>>>
>>>>
>>>> Thanks for considering this,
>>>>
>>> CLFSWM is under the GPL licence for historical and personal reason
>>> (mainly, I do not care of proprietary software).
>>> It is inspired by Stumpwm which is under the GPL licence. I've used few
>>> code from it and from Eclipse. This code is relatively easy to find:
>>>
>>> $ egrep -nri "eclipse|stumpwm" *
>>> AUTHORS:23:Stumpwm: http://www.nongnu.org/stumpwm/
>>> contrib/clfswm:23:# Original code and idea: http://stumpwm.antidesktop.net/cgi-bin/wiki/SetUp
>>> README:4: and [2]Stumpwm. Many thanks to them).
>>> README:108:   2. http://www.nongnu.org/stumpwm/
>>> src/tools.lisp:569:(defun find-free-number (l)          ; stolen from stumpwm - thanks
>>> src/keysyms.lisp:3:;;  This file is part of stumpwm.
>>> src/keysyms.lisp:5:;; stumpwm is free software; you can redistribute it and/or modify
>>> src/keysyms.lisp:10:;; stumpwm is distributed in the hope that it will be useful,
>>> src/version.lisp:4:;;  Borrowed from Stumpwm
>>> src/xlib-util.lisp:174:;;; Workaround for pixmap error taken from STUMPWM - thanks:
>>> src/xlib-util.lisp:178:;; in stumpwm. So far the only slot that seems to be affected is
>>> src/xlib-util.lisp:397:;;; Stolen from Eclipse
>>> src/netwm-util.lisp:67:;;; Taken from stumpwm (thanks)
>>>
>>> There is also the contrib/clfswm script from Xavier Maillard which is
>>> also based on Stumpwm code and under GPL licence.
>>>
>>> All that said, I'm really attached to the GPL licence even for Common
>>> Lisp code (BTW I've placed another Common Lisp project under the LLGPL
>>> licence).
>>> For now, I do not want to change the CLFSWM licence. So if you want to
>>> contribute to CLFSWM, you'll have to place your code under the GPL
>>> licence if you want to have it merged in the CLFSWM core code.
>>> Another solution is to make your hack in the contrib/ directory and
>>> place it in the licence of your choice.
>>> Also, CLFSWM (and Stumpwm I suppose) are placed under the GPL licence
>>> because they're expected to be run in free Common Lisp implementation
>>> (I regularly test CLFSWM in many of them) but not in closed source
>>> Common Lisp implementation.
>>>
>>> Indeed this is not a closed response and we have to discuss this on the
>>> clfswm-devel mailing list (in CC) with others. It'll be sad to loose a
>>> contributor.
>>>
>>> Regards,
>>>
>>> Philippe
>>>
>>>
>>>> Aad Versteden
>>>>
>>>>
>>>> PS: if you'd like to chat about it, feel free to add me on google
>>>> talk, or hop by on irc.freenode.net and send me, madnificent, a /msg
>>>> :)
>>>>
>>> Hmm, I'm not a lawyer and I'm not sufficiently fluent in English to
>>> discuss this question on IRC in a comfortable manner. Please continue
>>> this discussion on the clfswm-devel mailing list.