You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I'm not sure about the entire history of the license, but it looks to have been chosen when the project originally was hosted on the zope.org community site.
As the project as grown, I think its time to update the license to one that provides the same rights and protections but its more commonly known BSD 3-clause
As very minimum I would like to update the license from the current zope 2.0 to its more friendly written revision zope 2.1
The license proposed is MIT License.
MIT license - list of contributors and status on approval:
Hmm, the current copyright (attributing the library to the zope project) notice is a bit tricky then, isn't it? We should at least for new contributions have something like "the Python.NET project" instead.
it seems that back in the day (~2001) zope required any project hosted by it to have it's license. Reading through their pages the intent of the zope license was for it to apply for any software extensions to the zope software.
I think it's safe to say that python.net isn't particularly an extension to zope 😃
yup, sadly I'm familiar with the process. Thats why I offered updating to zope 2.1 as an alternative since it atleast is a step in the right direction.
note that my gripe with the zope 2.0 license is this line
This software is Copyright (c) Zope Corporation (tm) and Contributors. All rights reserved.
which was updated to
A copyright notice accompanies this license document that identifies the copyright holders.
and doesn't require Zope to be named on it as state by
The 2.1 version of the license is reusable. It supports having a consistent license for Zope and third-party products without requiring 3rd-party developers to assign copyright to Zope Corporation.
Zope 2.0 can give the wrong impression about the open-sourceness of the project
Yeah, I'm with you on all of that, but "officially" the copyright lies with the zope project currently, doesn't it? So we would need their approval for a licence change.
Since the 2.0 version doesn't specify an upgrade path, it shouldn't matter much to which new licence we switch, does it?
Note that Zope Corporation (http://www.zope.com/) looks shut down since 2015. Zope Foundation (http://www.zope.org) may still be active. If we spend the effort to upgrade by getting permission
from past contributors and Zope, then we may just go with more adopted
licenses such as BSD simplified or MIT?! It took me about an hour to
analyze ZPL when first introduced to pythonnet for legal reasons.
On Monday, June 27, 2016, Benedikt Reinartz [email protected]
wrote:
Yeah, I'm with you on all of that, but "officially" the copyright lies
with the zope project currently, doesn't it? So we would need their
approval for a licence change.
Since the 2.0 version doesn't specify an upgrade path, it shouldn't matter
much to which new licence we switch, does it?
@vmuriart@matthid@filmor I see you commented above, but can you please confirm your agreement on MIT license here? Others - sorry for the noise, but we need to get approval from everyone. Feel free to un-subscribe from this thread.
Do we know what changes were added by @tiran? We could always fork from before his additions and them re-patch with everyone else's changes if his weren't significant.
It's a huge headache though...
(You could also clean-room his changes after the fact if they were important.)
@bltribble A quick blame shows that none of his changes to the actual code are still "at the surface". The only ones left are changes to text files (TODO.txt, VS_README.txt, changes.txt) and entries for old python versions in the Python.Runtime.dll.config. The changes.txt (heavily outdated, should likely just be removed) contains the following with his tag:
First work on Python 2.5 compatibility. The destination version can be
set by defining PYTHON24 or PYTHON25. Python 2.6 compatibility is in
work.
Added VS 2005 solution and project files including a UnitTest configuration which runs the unit test suite.
Enhanced unit test suite. All test cases are combined in a single test suite now.
Fixed bugs in generics support for all Python versions.
Fixed exception bugs for Python 2.5+. When compiled for Python 2.5+ all managed exceptions are based on Python's exceptions.Exception class.
Added deprecation warnings for importing from CLR.* and the CLR module.
Implemented support for methods with variable arguments spam(params object[] egg)
Fixed Mono support by adding a custom marshaler for UCS-4 unicode, fixing a some ref counter bugs and creating a new makefile.mono.
Added a standard python extension to load the clr environment. The src/monoclr/ directory contains additional sample code like a Python binary linked against libpython2.x.so and some example code how to embed Mono and PythonNet in a C application.
Added yet another python prompt. This time it's a C application that embedds both Python and Mono. It may be useful as an example app for others and I need it to debug a nasty bug.
Implemented ModuleFunctionAttribute and added ForbidPythonThreadsAttribute.
The latter is required for module functions which invoke Python methods.
Added clr.setPreload(), clr.getPreload(), clr.AddReference("assembly name"),
clr.FindAssembly("name") and clr.ListAssemblies(verbose). Automatic
preloading can be enabled with clr.setPreload/True). Preloading is
automatically enabled for interactive Python shells and disabled in all
other cases.
New Makefile that works for Windows and Mono and autodetects the Python
version and UCS 2/4 setting.
Added code for Python 2.3. PythonNet can be build for Python 2.3 again
but it is not fully supported.
Changed the PythonException.Message value so it displays the name of
the exception class ("Exception") instead of its representation
("<type 'exceptions.Exception'>").
@filmor blame might be skewed and hide alot of the surviving changes he made since there's a commit in which the code was formatted to get rid of all those pesky whitespace/formatting issues we were having.
I'm well aware of that, just wanted to find out, whether all of this tiran-poking was even worth the effort as his last contribution is from 8 years ago or so.
Some of the features have been removed already or only relate to the build system, so we have a realistic chance of solving this ;)
if any of contributors needs a license to full JetBrains Product suite (Resharper, PyCharm, Rider, CLion, Intellij Idea), which can only be used for non-commercial development, then please let me know here or at my email address at denis(dot)akhiyarov(at)gmail(dot)com
We have total of 15 licenses on first come first serve basis until filled out.
In 2003, Free Software lawyers consulted for Mozilla's relicensing project and stated that relicensing with the permission of just 95% of contributors was fine, as long as there were no objections in the remaining 5%.
We're at 31/32 approved (96.8%) and had no objections from @tiran in the time we've tried to contact him. Looks like we are good to go with the re-licensing without having to rewrite @tiran's contributions.
Activity
filmor commentedon Jun 25, 2016
Hmm, the current copyright (attributing the library to the zope project) notice is a bit tricky then, isn't it? We should at least for new contributions have something like "the Python.NET project" instead.
vmuriart commentedon Jun 25, 2016
it seems that back in the day (~2001) zope required any project hosted by it to have it's license. Reading through their pages the intent of the zope license was for it to apply for any software extensions to the
zopesoftware.I think it's safe to say that python.net isn't particularly an extension to
zope😃den-run-ai commentedon Jun 25, 2016
re-licensing normally requires permission from all past contributors:
jelmer/dulwich#153
vmuriart commentedon Jun 25, 2016
yup, sadly I'm familiar with the process. Thats why I offered updating to
zope 2.1as an alternative since it atleast is a step in the right direction.vmuriart commentedon Jun 25, 2016
note that my gripe with the zope 2.0 license is this line
which was updated to
and doesn't require Zope to be named on it as state by
Zope 2.0 can give the wrong impression about the open-sourceness of the project
filmor commentedon Jun 27, 2016
Yeah, I'm with you on all of that, but "officially" the copyright lies with the zope project currently, doesn't it? So we would need their approval for a licence change.
Since the 2.0 version doesn't specify an upgrade path, it shouldn't matter much to which new licence we switch, does it?
den-run-ai commentedon Jun 27, 2016
Note that Zope Corporation (http://www.zope.com/) looks shut down since 2015. Zope Foundation (http://www.zope.org) may still be active. If we spend the effort to upgrade by getting permission
from past contributors and Zope, then we may just go with more adopted
licenses such as BSD simplified or MIT?! It took me about an hour to
analyze ZPL when first introduced to pythonnet for legal reasons.
On Monday, June 27, 2016, Benedikt Reinartz [email protected]
wrote:
vmuriart commentedon Jun 27, 2016
I had the same experience/concern when first starting to use
pythonnetmatthid commentedon Jun 27, 2016
I can +1 this. A switch to Apache 2 or MIT makes a lot of sense considering the latest developments in the .NET community...
den-run-ai commentedon Jul 29, 2016
I'm +1 on MIT license, it is top on github and openhub, also most permissive:
https://github.com/blog/1964-open-source-license-usage-on-github-com
https://www.blackducksoftware.com/top-open-source-licenses
http://choosealicense.com/licenses/
do @BartonCline @brianlloyd @davidanthoff @filmor @matthid @tonyroberts @vmuriart all agree before we ask permission from previous contributors?
brianlloyd commentedon Jul 29, 2016
I'd be fine with MIT.
davidanthoff commentedon Jul 29, 2016
Same here, not that there are many commits from my side in any case, I'd also be fine with MIT.
tonyroberts commentedon Jul 30, 2016
Fine with me
den-run-ai commentedon Aug 1, 2016
@vmuriart @matthid @filmor I see you commented above, but can you please confirm your agreement on MIT license here? Others - sorry for the noise, but we need to get approval from everyone. Feel free to un-subscribe from this thread.
I'm updating checklist on the top of this issue:
#234 (comment)
lstratman commentedon Aug 1, 2016
Sounds good to me.
33 remaining items
freakboy3742 commentedon Oct 17, 2016
@denfromufa I've just poked him as well.
den-run-ai commentedon Oct 18, 2016
@freakboy3742 thank you for this! I guess you meant that you poked @tiran?
Let me introduce 3 more contributors since August, who also agreed to MIT license:
@ArvidJB
@T3476
@dmitriyse
To all contributors, please reply to my offer about advanced Python book for pythonnet, if you are interested:
https://mail.python.org/pipermail/pythondotnet/2016-October/001828.html
bltribble commentedon Oct 18, 2016
Do we know what changes were added by @tiran? We could always fork from before his additions and them re-patch with everyone else's changes if his weren't significant.
It's a huge headache though...
(You could also clean-room his changes after the fact if they were important.)
freakboy3742 commentedon Oct 18, 2016
@denfromufa Correct - I poked @tiran. I may know some people that can perform a poke on his physical person, too... I'll see what I can arrange :-)
filmor commentedon Oct 19, 2016
@bltribble A quick blame shows that none of his changes to the actual code are still "at the surface". The only ones left are changes to text files (
TODO.txt,VS_README.txt,changes.txt) and entries for old python versions in thePython.Runtime.dll.config. Thechanges.txt(heavily outdated, should likely just be removed) contains the following with his tag:set by defining PYTHON24 or PYTHON25. Python 2.6 compatibility is in
work.
spam(params object[] egg)The latter is required for module functions which invoke Python methods.
clr.FindAssembly("name") and clr.ListAssemblies(verbose). Automatic
preloading can be enabled with clr.setPreload/True). Preloading is
automatically enabled for interactive Python shells and disabled in all
other cases.
version and UCS 2/4 setting.
but it is not fully supported.
the exception class ("Exception") instead of its representation
("<type 'exceptions.Exception'>").
den-run-ai commentedon Oct 19, 2016
@freakboy3742 please don't poke him too hard :)
swinstanley commentedon Oct 19, 2016
In the interests of a sensible rate of future development, is there some way I can transfer all rights and claims I might have to Harambe the gorilla?
bltribble commentedon Oct 19, 2016
@filmor That seems far too deep to attempt an exorcism...
vmuriart commentedon Oct 19, 2016
@filmor blame might be skewed and hide alot of the surviving changes he made since there's a commit in which the code was formatted to get rid of all those pesky whitespace/formatting issues we were having.
filmor commentedon Oct 19, 2016
I'm well aware of that, just wanted to find out, whether all of this tiran-poking was even worth the effort as his last contribution is from 8 years ago or so.
Some of the features have been removed already or only relate to the build system, so we have a realistic chance of solving this ;)
den-run-ai commentedon Nov 9, 2016
if any of contributors needs a license to full JetBrains Product suite (Resharper, PyCharm, Rider, CLion, Intellij Idea), which can only be used for non-commercial development, then please let me know here or at my email address at denis(dot)akhiyarov(at)gmail(dot)com
We have total of 15 licenses on first come first serve basis until filled out.
vmuriart commentedon Dec 4, 2016
@freakboy3742 did you have any luck daisy-chain poking @tiran ?
freakboy3742 commentedon Dec 5, 2016
@vmuriart Unfortunately not. I'll try again.
vmuriart commentedon Jan 8, 2017
I did some research on precedents and found:
We're at 31/32 approved (96.8%) and had no objections from @tiran in the time we've tried to contact him. Looks like we are good to go with the re-licensing without having to rewrite @tiran's contributions.