<font><span style="line-height:normal;background-color:rgba(255,255,255,0)">We really need to make a call here, we start to have a sizeable amount of code and the first release is near. I tend to think gplv2 is not an option because of the apache incompatibility. I would go for Apache if we want to avoid issues with anti-tivoization, otherwise gplv3.</span></font><br>
<br>To point out a concrete problem we could have with gpl3... My understanding is that you could not ship an activity based on sugar-web in the apple store, at least including the lib locally. I suppose it would be fine if you loaded it from a server, but then you need security restrictions if you implement any kind of system integration.<span></span><br>
<br>On Friday, 3 May 2013, Daniel Narvaez wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div dir="ltr"><div>Hello,<br><br>we need to decide how to license the new javascript libraries. I am mostly clueless about the topic and I'm honestly scared to start this thread, please be gentle :)<br>
<br></div>Following is the rationale I came up with for Agora. I think it probably applies to the sugar-html libraries too. Feedback would be very welcome as we are no expert.<br>
<br>---<br><br>I spent some time trying to decide which license is better for the
various part of Agora. It's an hard and important decision, I'm not a
lawyer and not even an expert but we need to make a call. My
understanding is that a license is better than nothing.<br><br>(L)GPLv2<br><br>* Copyleft. Requires all the modifications to be made freely available.<br>*
Incompatible with Apache. Pretty bad, a lot of code already licensed
that way and growing fast (especially in the javascript world).<br><br>(L)GPLv3<br><br>* Copyleft<br>* Compatiible with Apache.<br>*
Anti-tivoization clause. Mixed bag, would it prevent us to run on
hardware we are interested in? One problematic case I can think of is
distributing an activity through the Apple store. We wouldn't be able to
do that. Though people could still install the activity as a web app,
from the browser. Maybe that's good enough?<br>* Latest version. Better wording etc. Patents protection.<br>* We can distribute the sugar icons under LGPLv3, without requiring any relicensing, because of the "or later" clause.<br>
*
My understanding is that if xi-* is LGPL, proprietary applications
could still use it without making modifications. The situation is not as
clear as for the traditional linked libraries case but from <a href="http://www.gnu.org/licenses/lgpl-java.html" target="_blank">http://www.gnu.org/licenses/lgpl-java.html</a> I'd think we are fine.<br><br>Apache<br>
<br>*
Non copyleft. It would be more friendly to companies that might want to
reuse code in their products. But is that likely to happen? Both xi and
omega are pretty agora specific. Still I think it's a good license to
use for more generic bits that we might develop (I used it for some
python helpers I'm using in eta for example).<br>* It seems to be the best permissive license because of the patents protection. It's the most popular at least.<br><br>So I think there two choices basically:<br><br>
1 Copyleft VS non copyleft. I think copyleft has advantages and practically no real disadvantages for eta, xi and omega.<br><br>2
GPLv2 VS GPLv3. Compatibility with Apache would be good (maybe not
essential though? We could still use apache libraries I would think,
just not freely cut/paste code). Anti-tivoization is tricky, I honestly
can't make strong points one way or another. While I was initially
sympathetic with the claims that v3 is political I think <a href="http://tieguy.org/blog/2007/06/28/gpl-v3-the-qa-part-4-odds-and-ends/" target="_blank">http://tieguy.org/blog/2007/06/28/gpl-v3-the-qa-part-4-odds-and-ends/</a>
is a good rebuttal of that argument. I'm somewhat worried about not
being able to distribute on some devices but, especially since we can
always run remotely, I'm not convinced we should opt out of v3 because
of that.,<br clear="all"><div><div><br>-- <br>Daniel Narvaez<br>
</div></div></div>
</blockquote><br><br>-- <br>Daniel Narvaez<br><br>