May be of interest: the India open data community (datameet.org) just did a relicensing of many of the popular map datasets like boundaries and constituencies from cc-by-sa 2.5 → cc-by 4.0 License data with `CC-by 4.0` to be compatible with Government Open Data License · Issue #56 · datameet/maps · GitHub .
When a sharing practice yields to less sharing practice, it is not good news for those who like to sustain sharing. A license that was initially devised to promote more and more sharing will be considered as a restriction by only those who do not want to encourage data sharing in all situations. Yes, the copyleft clause does restrict, but restrict precisely those practices that have the intent to stop sharing.
The use of both CC BY and CC BY-SA for map datasets continues to expose the fundamental misunderstanding people have about copyright.
You cannot copyright data. You can copyright the maps (paper or digital) because they are an expression of the data, but the data themselves cannot be copyrighted.
You should not copyright data. Having said above that you cannot copyright data, I am following with you should not copyright data because, in their ignorance (or misunderstanding), people continue to apply copyright licenses to data. As a result, potential users, also in their ignorance of misunderstanding, continue to accept the licensing of such data. This just perpetuates the ignorance.
This is precisely why CC0 was invented (in 2008), but after 12 years, this battle seems like a lost cause. In that sense, perhaps CC did more harm than good as it caused the proliferation of this misunderstanding. On the other hand, if CC BY encourages people to both share, use and give credit for the data, then that is only a good thing, and the aforementioned harm can be forgiven.
sharing can never lead to less sharing. The original that is shared continues to exist and can continue to be shared
I misspoke because obviously you can copyright data… after all, that is precisely what the folks mentioned in OP did. But doing so is wrong and misleading. You should not copyright data. Here is something relevant I wrote a while back… it is about (developing one’s own) licenses but is just as applicable to applying licenses
but restrict precisely those practices that have the intent to stop sharing
BY SA certainly restricts those practices that have the intent to stop sharing like the creation of a copyright derivative work.
A side effect of this is that it does not recognize that sharing could happen in more than one way.
A real example is the data produced by the largest free map data project - OpenStreetMap. All the data is licensed under ODbL which is a similar BY SA license for a database. It so happens that CC BY SA data is incompatible to be imported into an ODbL database!
After carful examination, the OSM license working group declared that even CC BY content is not fully compatible, and needs an explicit waiver from the original author.
Creating multiple silos of share alike content incompatible with each seems to be an unfortunate side affect of defining sharing too narrowly. What would be the solution in this case where a data license needs to be chosen to be compatible with other open projects of similar spirit?
Yes. But the derivatives may not be. Therefore when you use the term less/more, the question arises how can it lead to increase sharing when derivatives need not be shared. My simple understanding is that in the situation when both the original and the derivatives are shareable, we increase the benefits of sharing.
Do you have a counter-intuitive model that shows otherwise?
What assumptions are we making here, when one says “we should not copyright data”? People do copyright data is a different matter.
What comes to my mind is the belief that data is not an expression, and certainly not a creative expression, therefore data cannot be copyrighted. Please clarify if this is what is behind the idea that data should not be copyrighted.
Those who want to make derivatives of BY-SA content may be deterred from doing so because they may not want to apply BY-SA (for various reasons – it is not a straightforward decision-making process for everyone). In such cases, they may start from scratch and develop a completely proprietary stack, or use some alternative that is not BY-SA. That means your SA content will not be used as much as it could have been used. However, if someone uses my BY work and makes the derivative proprietary, my work continues to exist as BY.
By using BY licensing, I never decrease the amount my work could be shared. My job was creating my work and releasing it under BY. My work is done, my duty is done. It is not my responsibility to police what license someone else uses. Freedom should be voluntary, not imposed.
there are no assumptions… it is a matter of law. Facts/data cannot be copyrighted, only expressions of such facts/data can be copyrighted. Copyright depends on the presence of creativity. Data are not creative. They just are.
That people copyright data is wrong. No one challenges them in the court, but if they did, the licensors would probably lose. Why does no one challenge them? Because usually it is not worth the trouble. But sharing does get slowed with copyright licenses. I might just decided that it is not worth using something that is copyrighted even if it shouldn’t have been copyrighted in the first place.
As another user in this thread @planemad mentions, OSM ended up pushing for the creation of and using ODbL. We at CC wanted them to use CC0 but OSM was determined to ensure attribution. In our view, that was a wrong call, but it was their prerogative. They did it. But doing so does restrict usage… for example, you can’t mix a SA-type license with any other license.
Oh, all this is super-boring. That is why almost a decade ago I decided to leave this nonsense and just give away everything under CC0. It is so much nicer when I create something but choose not to own the rights to it. See https://punkish.org/Freedoms-Just-Another-Word
Freedom needs legislative protection. Legislation is to grant freedom for everyone, and not to some. Therefore, granting freedom not to grant freedom to others is actually a vulnerable model of ensuring freedom.
Until that time when all creative expressions are available in public domain, legislative protection of freedom is required. In a heterogeneous society, we need to agree on protocols/rules to grant creativity to all. That I understand is the intent of SA. Calling this intent as a restriction because it does not give choice to those who do not want to make creative expression equitable, sustains exploitation and not freedom.
The need for protecting/policing a common/public space arises because everyone’s intent is not to sustain the public space. We can’t impose freedom, we can protect freedom.
I elaborated my view on why this assumption is not tenable in another post.
@GN that sentence above is mixing different thoughts. Sure, freedom needs legislative protection, which is what the license does. The license (whichever license one uses) gives the licensor the power to grant the licensee permissions for the use of one’s content that the licensor wishes to grant. Beyond that seems like overreach. At no time is the licensor’s content endangered. It continues to exist as it was, compliant fully with the wishes and whims of the licensor. These goods are non-subtractable. By giving you the right to use my digital work as you see wish, I am in no way diminishing the value and power of my work. If someone else wants to take my work, build upon it, and wants to encumber the original work that s/he created, well, that is their wish. Why should I impose upon them my wishes?
Anyway, this thread has no satisfactory ending because it is based in personal philosophy. I believe that my wishes were fulfilled in the creating of whatever it is that I set out to create. I am happy with that and want nothing more. I don’t even want attribution, and I certainly don’t want to impose my wishes on someone else. In that spirit, I don’t even care what license someone else uses. If someone else wants to use a SA-type license, all strength to them. I will choose, at my convenience, to use their work or not use it.
While as a principle i agree, once we have started to codify freedom in a legal manner, it looks like we’ve ended up in a situation which has given birth to multiple definitions of freedoms not compatible with each other.
Will once again bring up the real scenario of two licenses - CC BY SA and ODbL which embodies the exact same spirit to guarantee perpetual freedom through transformations but has ended up being incompatible to mix with each other.
The result is that one needs to choose between the different the worlds of freedom one wishes to inhabit, and that is going to be lesser than the sum of all the free worlds.
PS: an interesting background with the OSM project which uses ODbL license is that it was originally under the CC BY SA license but was deemed to be impractical for a database and the project was relicensed in 2011.
While much of the discussion here is on a higher level of the right approach, id also like to share my feelings as a content author with substantial contributions on 3 open knowledge projects over 10 years:
- OpenStreetMap ODbL
- Wikipedia: CC BY SA
- Wikidata CC 0
Personally, i feel most rewarded with my contributions to Wikidata. Unlike Wikipedia and OSM which will continue to thrive in its own niche worlds, i feel Wikidata will permeate all worlds freely and nobody can stop it. This makes me feel part of a really powerful and free organism and nothing at all can take away or erode that feeling.
To echo sentiments of @punkish, to me it honestly does not matter what someone else does with my work, every reuse is my reward and it in no way subtracts from my efforts.
There’s absolutely no reason though that everyone should feel the same and it is totally respectable if one only feels rewarded with a certain kind of reuse. It comes down to the personal motivations of the creators and we should be fine with everyone having the freedom to exercise their tastes.