Hi,
The OpenLayers package (already in Fedora, owned by Cristian in CC and of which I'm a co-maintainer) bundles a JavaScript minimizer called jsmin [1]. I know this is bad practice to have a tool bundled like this, and I wanted to remove it from the source RPM and make it use a system version of jsmin that would be installed in its own package.
However, I saw that jsmin had already been submitted to Fedora and was refused because the license specifies that « The Software shall be used for Good, not Evil. » [2]
I discussed it with the OpenLayers devs (see attached IRC log) and it seems we have 2 possibilities.
1. The jsmin.py script was rewritten from scratch to mimic the behavior of the C original version. It currently contains the following license header:
==== # This code is original from jsmin by Douglas Crockford, it was translated to # Python by Baruch Even. The original code had the following copyright and # license. # # /* jsmin.c # 2007-01-08 # # Copyright (c) 2002 Douglas Crockford (www.crockford.com) # # Permission is hereby granted, free of charge, to any person obtaining a copy of # this software and associated documentation files (the "Software"), to deal in # the Software without restriction, including without limitation the rights to # use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies # of the Software, and to permit persons to whom the Software is furnished to do # so, subject to the following conditions: # # The above copyright notice and this permission notice shall be included in all # copies or substantial portions of the Software. # # The Software shall be used for Good, not Evil. # # THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR # IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, # FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE # AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER # LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, # OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE # SOFTWARE. # */ ====
The part between JavaScript comments is the original problematic license (starts with « /* » and ends with « */ »).
My question is, as this seems to be a « clean room implementation », could it have a different license than the original jsmin ? This would make the python version suitable for Fedora (correct me if I'm wrong) and that would allow me to build OpenLayers using it.
2. The second possibility is to simply build OpenLayers without jsmin. This would however result in a much bigger JavaScript file, leading to worse performances.
That's a solution I'd rather avoid, but if this is the only possibility, I'll do it.
Best regards,
[1] http://www.crockford.com/javascript/jsmin.html [2] https://bugzilla.redhat.com/show_bug.cgi?id=455507
----------
Mathieu Bridon (bochecha)
Hi,
On Sun, Aug 2, 2009 at 21:00, Mathieu Bridon (bochecha)bochecha@fedoraproject.org wrote:
Hi,
The OpenLayers package (already in Fedora, owned by Cristian in CC and of which I'm a co-maintainer) bundles a JavaScript minimizer called jsmin [1]. I know this is bad practice to have a tool bundled like this, and I wanted to remove it from the source RPM and make it use a system version of jsmin that would be installed in its own package.
However, I saw that jsmin had already been submitted to Fedora and was refused because the license specifies that « The Software shall be used for Good, not Evil. » [2]
I discussed it with the OpenLayers devs (see attached IRC log) and it seems we have 2 possibilities.
- The jsmin.py script was rewritten from scratch to mimic the
behavior of the C original version. It currently contains the following license header:
[snip]
My question is, as this seems to be a « clean room implementation », could it have a different license than the original jsmin ? This would make the python version suitable for Fedora (correct me if I'm wrong) and that would allow me to build OpenLayers using it.
Any thoughts on this solution? Is that legally possible?
If so, I'll try to contact the writer of the Python implementation to see if he is willing to relicense his script.
- The second possibility is to simply build OpenLayers without jsmin.
This would however result in a much bigger JavaScript file, leading to worse performances.
That's a solution I'd rather avoid, but if this is the only possibility, I'll do it.
Best regards,
[1] http://www.crockford.com/javascript/jsmin.html [2] https://bugzilla.redhat.com/show_bug.cgi?id=455507
----------
Mathieu Bridon (bochecha)
On 08/08/2009 01:20 PM, Mathieu Bridon (bochecha) wrote:
Any thoughts on this solution? Is that legally possible?
If so, I'll try to contact the writer of the Python implementation to see if he is willing to relicense his script.
Sorry for the delay here, not sure how this fell off my radar.
So, it is legally possible to do a clean room reimplementation of this code, but it does not look like that is what happened with the existing jsmin.py.
Basically, what someone would have to do is to write a version of jsmin, looking only at an "algorithm description", but never jsmin.c.
Red Hat Legal compared this jsmin.py to the jsmin.c code, and it is their opinion that it is not a clean-room reimplementation, but rather a "conscious" translation from C to Python.
If I had to guess, if pressed, the jsmin.py author will admit to having looked at jsmin.c.
So, you can either find someone to make a jsmin in a clean room reimplementation, then use it, or rework your package to not use it.
~spot
So, it is legally possible to do a clean room reimplementation of this code, but it does not look like that is what happened with the existing jsmin.py.
Basically, what someone would have to do is to write a version of jsmin, looking only at an "algorithm description", but never jsmin.c.
Ok, thanks for confirming this was a possibility.
Red Hat Legal compared this jsmin.py to the jsmin.c code, and it is their opinion that it is not a clean-room reimplementation, but rather a "conscious" translation from C to Python.
If I had to guess, if pressed, the jsmin.py author will admit to having looked at jsmin.c.
I actually asked him, to be sure, and that's what he told me as well, that he simply translated into Python.
So, you can either find someone to make a jsmin in a clean room reimplementation, then use it, or rework your package to not use it.
I removed the jsmin tools from the source tarball of OpenLayers and rebuilt it for F-11 and Rawhide (I don't have the commit access in F-10, and it will be EOL-ed soon anyway, if that really matters I'll ask for the commit ACL and do it there as well).
In the future, I'll try to see if another Javascript minifier couldn't be used instead of jsmin and what the OpenLayers devs think about it.
Best regards,
----------
Mathieu Bridon (bochecha)