RFD: MIT/Lua License for Libmarpa

[Ruslan Shvedov]

Further to the talk on IRC [1] with Jeffrey Kegler, the author of Libmarpa, on his request, I'm restarting the discussion [2] about switching Libmarpa to the MIT/Lua license [3] so that to ask opinions on the matter and make sure that we have consensus on that licensing change. 

As detailed in [2], this licensing change will not apply to the code derived from LGPL'd code written by others, such as obstacks and AVL.

[1] http://irclog.perlgeek.de/marpa/2015-02-12#i_10104254
[2] http://groups.google.com/d/msg/marpa-parser/5IVTcnCqGwo/EfPbnbOOp6oJ
[3] http://www.lua.org/license.html

[Jeffrey Kegler]

Readers of this group may recall that conversion of Libmarpa to an MIT/Lua license was discussed, but got side-tracked because I combined the proposal with an unpopular proposal to change to copyright assignment.  Ruslan S. has generously agreed to take this on, so that I can concentrate on coding.

I believe that conversion to the MIT license had a consensus, but it is an important matter, and I wanted to be sure.

By the way, I am grateful to David Yingling for taking on the task of expressing the growing dislike of copyright assignment, and thereby stopping me from taking what might well have been a counter-productive step.

--
You received this message because you are subscribed to the Google Groups "marpa parser" group.
To unsubscribe from this group and stop receiving emails from it, send an email to marpa-parser...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[David Yingling]

I'm all for switching to MIT, because it's still open-source, and could help
make it more popular.

Doing so should be straightforward.
1. Enumerate all contributors to LibMarap (maybe Marpa::R2??? maybe not). Git
should make this a breeze. Mayben even post a list of everyone you think
is a contributor to this list, and see if anyone is missing.
* Do *not* forget possibly hidden contributors that lack credit in github.
It's up to Jeffery Keggler to note if any contributors' code/other
contributions may have been committed to git under his name. Perhaps
directly from email or this email list? Or was he a stickler for github
pull requests, and, therefore, we don't have to worry about this? I don't
know, but its worth asking.
2. Email them all asking them to relicense their commits under MIT. Include the
exact text of the license you want them to agree to switch to.
* Is just a reply to an email "enough" legally speaking?
3. Commit "their permission" to git some how.
* Just copy and paste the emails that were sent to them, and their replies
to the new Lua-MIT LICENSE file? Is that enough?

An alternative way would be to ask contributors to "recontribute" their
contributions by adding comments to the code in question where they agree to
relicense their contribution to Lua-MIT. Maybe, provide a sentence or 2 of text
for them to copy and paste and use where they fill in their name. This way
would probably be stronger in court if it ever would come to that.

Any other issues?

On Thu, 12 Feb 2015 08:27:23 -0800
Jeffrey Kegler <jeffre...@jeffreykegler.com> wrote:

> Readers of this group may recall that conversion of Libmarpa to an MIT/Lua
> license was discussed, but got side-tracked because I combined the proposal
> with an unpopular proposal to change to copyright assignment. Ruslan S.
> has generously agreed to take this on, so that I can concentrate on coding.
>
> I believe that conversion to the MIT license had a consensus, but it is an
> important matter, and I wanted to be sure.
>
> By the way, I am grateful to David Yingling for taking on the task of
> expressing the growing dislike of copyright assignment, and thereby
> stopping me from taking what might well have been a counter-productive step.
>

Thanks. I have some crazy ideas for using Marpa in the future that might
happen or might turn into nothing, but I don't want my annoying hatred for
copyright assignment making the decision for me. I think, in academia
copyright assignment flourishes, because scientists want their work shared with
others and published in well-respected journals, so they're willing to
do what it takes to do this, but in FOSS a different set of issues is at work
with freedom, and what exactly the sort of freedom you mean (GPL-style freedom
or BSD-style freedom?) plays a big role that copyright assignment impedes.

[Ron Savage]

A scary article on algorithmic patenting: http://yro-beta.slashdot.org/story/15/02/17/2032207/algorithmic-patenting

[Durand Jean-Damien]

Frightening -; !