[m-dev.] [m-users.] Closed source Mercury projects on Windows

[Mark Brown]

On Sun, May 27, 2018 at 9:05 PM, Julien Fischer <jfischer at opturion.com> wrote:
>
> Hi all,
>
> On Sun, 27 May 2018, Zoltan Somogyi wrote:
>
>> On Sun, 27 May 2018 19:57:50 +1000, Mark Brown <mark at mercurylang.org>
>> wrote:
>>>
>>> Yes, I would object, as I think there is a better way to achieve to
>>> intended effect. Adding a prelude or making other changes is tricky if
>>> you consider that the document uses negation heavily and is (of
>>> course!) recursively defined: does "this License", in 2(c) for
>>> example, refer to the original or the modified version?
>>>
>>> I would like to propose that we release the standard library under
>>> something like the following license, in addition to the LGPL: "You
>>> may, without restriction, copy and distribute the Library (or a
>>> portion or derivative of it) in object code or executable form as
>>> produced by the Compiler."
>>>
>>> I believe this permits only what was intended all along, namely that
>>> you should be able to write and distribute Mercury programs and
>>> libraries without making them open source. Does anyone disagree that
>>> it would have this effect?
>>
>>
>> No disagreement here. That would work as well, and even better,
>> for the reasons you cite.
>>
>> Do you have a proposal for an actual license text, or a model for us to
>> follow?
>
>
> The wxWindows library license is one such example of a static linking
> exception.  I suggest looking at modified LGPLs that have been OSI
> approved, rather than just making ad-hoc modifications to the LGPL.
> (Such ad-hoc modifications cause all sort of trouble.)

The wxWindows license looks right to me. Clause 2 is what I was trying
to get at with the above.

>
> My own feeling is that rather than using a modified LGPL, it would be
> more straightforward to relicense the runtime, stdlib etc under the
> BSD or MIT license.

How would that be more straightforward than relicensing it under the
wxWindows license?

I think the wxWindows license reflects what contributors - going back
a long way, mind you - actually intended from the library license, so
I'm pretty comfortable changing to that to address the technical
issues. But I don't see that the same can be said of BSD or MIT.

Mark