[License-discuss] Proposed license decision process
Nicholas Matthew Neft Weinstock
nweinsto at qti.qualcomm.com
Thu Dec 13 23:41:38 UTC 2018
We agree regarding the first two licenses, I'll cut those out for readability.
Regarding BSD, it seems like you're saying that you think it would (or should) not be accepted by OSI if it were newly proposed today. So would it be fair to say that taking up the new OSD would include some caveat that licenses already approved as "Open Source" would be grandfathered in even if they don't meet the new definition? That would certainly help to avoid confusion from licenses potentially being recategorized.
There are three open source licenses OSI has approved (RealNetworks, Reciprocal Public License, and Apple Public Source License) that expressly say "In consideration of, and as a condition to, the licenses granted to You under this License ..." Of course, these licenses aren't seen very frequently, but they could run afoul of a "no consideration" definition.
The ideas of conditions and consideration aren't always determined based on whether they're about money, and they're not always distinct. If a license says "You have a license to use my Trademark on the condition that you pay me $XYX" is the money consideration or a condition of the license? What if it says "You will have a license to use my Trademark as long as you continue to pay me $XYZ per month," does that change whether the money is consideration or a condition of the license?
To a more relevant example, does it matter if a license says "You have a license to create Derivative Works, subject to the following conditions" or "You have a license to create Derivative Works, subject to the following restrictions" or "You have a license to create Derivative Works, provided that you do the following"?
I could also see a difference whether something is a condition or consideration based on what the licensee is being asked to do (or not do). If it's a right you don't currently have, then telling you how you can exercise the rights being licensed is more of a condition. You don't have the right to make a Derivative Work without a license, so telling you that you can only make Derivative Works under license XYZ is a condition of receiving a license to make Derivative Works. But if it's a right you currently have, then telling you to give up that right in order to get the license is more like consideration. You currently have the right to make a Parody (as fair use), so if the license says you waive your right to make a Parody that would be consideration. (sorry, I couldn't think of a better example of this in the Open Source ecosystem)
From: Lawrence Rosen <lrosen at rosenlaw.com>
Sent: Thursday, December 13, 2018 1:23 PM
Subject: RE: [License-discuss] Proposed license decision process
Nick Weinstock wrote:
> To your question below, I can cite two examples of Richard's concern:
And you also cited two examples of your own concern about unapproved/un-approvable licenses. Thanks! I appreciate that.
* BSD: True. The patent license is not express. I have complained about that loudly every time someone proposes another BSD version. But our community is simply not worried about that. However, if a license now is proposed without an express patent grant, I'd object to it vociferously based on the definition of "open source software" that you quoted. OSI recently disapproved a license that expressly excluded a patent license, written that way purposefully to collect consideration. On the other hand, licenses from universities or research institutions may try to limit patent licenses based on previous contractual or legal requirements. That is why it becomes important to define "open source software" as software that is "actually distributed under terms that grant...," so that nobody can claim that their software is open source merely because they can see it. Is it "actually distributed" or "terms that grant" that concerns you? The W3C Royalty Free Patent Policy requires only that "the RF license conforming to the requirements in this policy shall be made available by the licensor as long as the Recommendation is in effect." The hope and expectation is that actual patent licenses won't be needed. That was also the approach taken by the Open Web Foundation. What is OSI's position on this?
* There are no open source copyright or royalty-free patent licenses that impose "consideration". There is some confusion in our field about the difference between "consideration" and "conditions" in licenses. OSI accepts license conditions that related to copyright or license enforcement - such as copyleft, attribution, trademark, the warranty of provenance, jurisdiction, patent defense - but those are not forms of consideration. For example, the copyleft "condition" for the licensee to reciprocate with his/her own software doesn't mean that anyone proposes to make money off that condition; copyleft licenses are granted for the purpose of creating "open source software," which is its own reward. Academic licenses, on the other hand, treat the "condition" of attribution as its own reward, even though there is no way to calculate the actual value of any such pleasure.
From: Nicholas Matthew Neft Weinstock <nweinsto at qti.qualcomm.com<mailto:nweinsto at qti.qualcomm.com>>
Sent: Thursday, December 13, 2018 11:37 AM
To: lrosen at rosenlaw.com<mailto:lrosen at rosenlaw.com>; license-discuss at lists.opensource.org<mailto:license-discuss at lists.opensource.org>
Subject: Re: [License-discuss] Proposed license decision process
This crossed in the ether with my response to Richard.
To your question below, I can cite two examples of Richard's concern:
* Ms-LPL is generally viewed as not "Open Source" because it has a platform limitation. It's not listed in SPDX or on OSI. It would satisfy this definition.
* Code Project Open License is sometimes viewed as not "Open Source" because it has a "fields of endeavor" limitation (may not be used for illegal, immoral, or improper purposes). It is listed in SPDX, but not on OSI. It would satisfy this definition.
I can also cite two examples of my concern, that licenses traditionally viewed as "Open Source" could be excluded by a highly literalist reading of the OSD:
* A highly literalist reading of "actually distributed under terms that grant" could suggest that the copyright and patent license terms must be express. The standard 3-clause BSD license does not make any mention of patents, and could thus fail the OSD.
* A highly literalist contemplation of "without payment of royalties or other consideration, to distribute the unmodified or modified software" could extend "other consideration" to actions that require the licensee to become a licensor, such as requiring binary distribution to also make the accompanying source (including the licensee's modifications) available under the same terms. Copyleft licenses such as GPL could thus fail the OSD.
Note: a highly literalist reading might also exclude CPOL, because it requires that a distributing licensee must ensure that recipients agree to the license, which could be another "other consideration."
From: Lawrence Rosen
Sent: Thursday, December 13, 2018 10:09 AM
Subject: Re: [License-discuss] Proposed license decision process
Richard Fontana wrote:
> I can easily come up with hypothetical licenses that would seem not to fail a highly literalist reading of the OSD, but which historically would never have been *treated* as conforming to the OSD, because of an obvious failure of the license to provide software freedom as traditionally understood in the community.
Can you please cite examples that we've screwed up (or create a hypothetical) because of a "highly literalist reading of the OSD"?
"Traditionally understood?" You sound like the late Justice Antonin Scalia! (Sorry; that crack is ad hominem!) :-)
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the License-discuss