What license does this software use?

[jamesan]

I'm assuming everything in this repository is open source/copy-left like GPLv2 or GPLv3, but it doesn't seem to indicate explicitly. Neither this repo, nor its predecessor, https://github.com/wertarbyte/autorandr, nor any references to autorandr or autodisper return any clues.

I'm updating Arch Linux's user contributed AUR package for autorandr and noticed its license field is blank. This would greatly help in my housekeeping of that package build. Thanks!

[phillipberndt]

The original author didn't give autorandr an OSS licence, see wertarbyte/autorandr#24, and has not replied to requests for over a year now. I'd guess that the other authors that contributed over the years would be willing to GPL the code, but without Stefan's consent, we AFAIK can't license autorandr unless we rewrite his code.

autorandr has four non-trivial parts: Storing configurations, loading configurations, fingerprinting and auto configuration. tachylatus has recently rewritten the part that is responsible for loading, and also most of the part that stores away configurations. Auto configuration was introduced by Timo Bingmann. The detection code (setup_fp) still is the one by Stefan.

What we could do is to rewrite the fingerprinting from scratch. IANAL, but I believe that the new author could then, together with Timo and Helge (tachylatus), license our version of autorandr. I'll CC them both, and also wertarbyte. He recently committed to another repository, so he might read this!

[phillipberndt]

@wertarbyte @tachylatus @bingmann What's your opinion on this?!

[bingmann]

I'm happy with whatever open-source license you're happy with.

[tachylatus]

I second bingmann's opinion :-)

[phillipberndt]

Let's go with the GPL then. I already reworked the fingerprinting, but after doing some research on the topic I believe that we need to have all authors' consent. Further contributors were @chrisdunder, @maciex, @stormc, @mrj10 and @formorer. Another script within this repository was written by @queezythegreat.

Do you consent with publishing your contributions to autorandr under the terms of the GPL 3?

[formorer]

On Fri, 02 Jan 2015, Phillip Berndt wrote:

Let's go with the GPL then. I already reworked the fingerprinting, but after doing some research on the topic I believe that we need to have all authors' consent. Further contributors were @chrisdunder, @maciex, @stormc, @mrj10 and @formorer. Another scripts within this repository was written by @queezythegreat.

Do you consent with publishing your contributions to autorandr under the terms of the GPL 3?
I would prefer MIT, but GPL is also fine.

Alex

[queezythegreat]

Hi, I think both MIT and GPL are fine, I have no objections. I also like the MPL License 2.0.

[stormc]

I'm fine with any license you find appropriate

[maciex]

I think GPL is fine.

[phillipberndt]

In bug #8, tachylatus suggested to rework the storage/application of configurations to account for pluggable graphics cards. My comment there might also be relevant to this issue:

Since this is a major change, let's wait a few days with this: Regarding the licensing issue #7, I noticed that much of wertarbyte's code is too trivial to be rewritten in another way, and I don't know enough about copyright law to be absolutely certain that this is not a problem. I wrote him an email yesterday, and hopefully he'll reply. If he doesn't, this conceptual change would be a great opportunity to rewrite the whole thing from scratch, keeping configuration/parameter compatibility of course. (autorandr isn't that big/complicated, after all.. and we could switch to Python or Perl to ditch the awk/sed scripts.)

The background is that in the meantime, I've read that such trivial code lines could still be a problem (there is an example with a copyrighted /bin/true on google). That what we work with here is a fork of wertarbyte's code obviously doesn't make that simpler. I don't believe that autorandr will ever be the object of a lawsuit, or that wertarbyte will sue anyone about it, but I also don't want to be the one to take the chance that I'm wrong. So if wertarbyte does not reply, we'll have multiple choices:

• Someone of you forks my license branch, optionally reverts bbe8f8b, and adds a copyright notice at the top. We'll all work with that code base, from then on.
• We switch to a true rewrite. This could either be another bash script (in which case we could reuse much, do a git blame on autorandr in the license branch), or be in another language, as written above. Here's a quick showcase of how this could be done in Python. The script should be fully (except the configuration file) compatible with autorandr by now (and it additionally supports transformation matrices, gamma levels, refresh rates & reflections), but still

  This is very early and experimental code, and I haven't tested it very much. [..] Please consider this as a suggestion on where we could go with autorandr rather than something I'd like to impose on anyone.. I'm open to suggestions.

  tachylatus started [a perl version](https://github.com/tachylatus/autorandr/blob/perl-experiment/autorandr.pl), if you'd prefer that. Or, if you'd prefer the rewrite option and to stick with bash, go ahead and do that.

I'm fine with any of these.

[ccdunder]

I prefer GPL.

~ cd
On Jan 7, 2015 10:10 AM, "Phillip Berndt" notifications@github.com wrote:

In bug #7 #7,
tachylatus suggested to rework the storage/application of configurations to
account for pluggable graphics cards. My comment there might also be
relevant to this issue:

Since this is a major change, let's wait a few days with this: Regarding
the licensing issue #7
#7, I noticed that
much of wertarbyte's code is too trivial to be rewritten in another way,
and I don't know enough about copyright law to be absolutely certain that
this is not a problem. I wrote him an email yesterday, and hopefully he'll
reply. If he doesn't, this conceptual change would be a great opportunity
to rewrite the whole thing from scratch, keeping configuration/parameter
compatibility of course. (autorandr isn't that big/complicated, after all..
and we could switch to Python or Perl to ditch the awk/sed scripts.)

The background is that in the meantime, I've read that such trivial code
lines could still be a problem (there is an example with a copyrighted
/bin/true on google). That what we work with here is a fork of
wertarbyte's code obviously doesn't make that simpler. I don't believe that
autorandr will ever be the object of a lawsuit, or that wertarbyte will sue
anyone about it, but I also don't want to be the one to take the chance
that I'm wrong. So if wertarbyte does not reply, we'll have multiple
choices:

• Someone of you forks my license branch
  https://github.com/phillipberndt/autorandr/tree/license, optionally
  reverts bbe8f8b
  bbe8f8b,
  and adds a copyright notice at the top. We'll all work with that code base,
  from then on.

• We switch to a true rewrite. This could either be another bash
  script (in which case we could reuse much, do a git blame on autorandr
  in the license branch), or be in another language, as written above. Here's
  a quick showcase of how this could be done in Python
  https://github.com/phillipberndt/autorandr/blob/license/autorandr.py.
  The script should be fully (except the configuration file) compatible with
  autorandr by now (and it additionally supports transformation matrices,
  gamma levels, refresh rates & reflections), but still

  This is very early and experimental code, and I haven't tested it very
  much. [..] Please consider this as a suggestion on where we could go with
  autorandr rather than something I'd like to impose on anyone.. I'm open to
  suggestions.

  tachylatus started a perl version
  https://github.com/tachylatus/autorandr/blob/perl-experiment/autorandr.pl,
  if you'd prefer that. Or, if you'd prefer the rewrite option and to stick
  with bash, go ahead and do that.

I'm fine with any of these.

—
Reply to this email directly or view it on GitHub
#7 (comment)
.

[phillipberndt]

wertarbyte just commented in wertarbyte/autorandr#24

After a while, I finally got around to this - licence should be GPLv3, so feel free to package :-)

Which means that autorandr is now officially GPLv3. For now, I've reverted the commit that removed parts of his code from the license branch and merged the branch into master. I'll still have to decide on how to move forward with the Python version. I will definitively maintain both for some time, but whether I do this in a single branch, two branches, repositories, under which name, etc. is all open. If you're interested in the future of both versions (e.g. if you maintain a package for this fork), subscribe to #12 for updates on that.