-
-
Notifications
You must be signed in to change notification settings - Fork 357
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Creative Commons text license handling #493
Comments
HappyLittleRat also had the idea of asking the CC folks for advice. |
I agree. That was part of the reason I tried to explain everything in sufficient depth. |
ShareAlikeI'm also 50/50 on ShareAlike. While it is a good tool to make sure, that every modification of the AAA and therefore the ideas and work of the AAA stay open, it might be a problem if, let's say, a teacher wants to use a text snippet of the AAA as part of a text in school, which they can't ShareAlike for some reason. The question is here: Does including a text snippet trigger the ShareAlike or is it just triggered, if you release a modified version? If it's the first case, I think we shouldn't use ShareAlike, if it's the second however, ShareAlike seems to be a good option. Global or Local LicenseI think we should license each chapter locally. By doing so, we keep the list of authors short and don't have one bloated contributors file. Minor ChangesIf I understand it correctly, each modification of the text would produce a new modified version of it, which requires attribution to the new author. While this is absolutely fine for text revisions and additions, this would also mean, that each fixed typo would result in a new contributor added to the authors. The Contribution ProcessAnother question is, how the contribution process would work with the license. Does stating that all texts are licensed under CC-BY-SA implicitly makes the author of a PR publishing their text / modification / typo under the license or do we need them to do this explicitly? |
Commercial UseWhile the NonCommercial license makes this project non-free (by FSF definition), it prevents big companies from selling and abusing the AAA. I think not using NonCommerical isn't too much of a problem, if the AAA would be at a size, that allows it to sell and ship books in an easy and efficient way. Since the AAA isn't at such a size at this point, companies have a clear advantage compared to the AAA and could easily make a lot of profit of printed versions. A solution could be, that text contributions get published under NC and an explicit right for commercial use is given to Leios or a potential AAA organisation. |
Share-Alike / Non-Commercial:I honestly think Share-Alike is predatory and shady, but I understand that it prevents people from taking the book, modifying the license, and then selling it. The Non-Commercial is much more straightforward and usable for this purpose. It just says, "Hey. You can do what you want with this, but you cannot sell it unless you are one of the people listed here." Share-Alike says, "You can use this, but only in the way we want you to." It boggles my mind that the FSF thinks the former is somehow less free than the latter, but I didn't draw the lines here. Again, I really don't care that much... It's just that Share-Alike bugs me a bit. I think the teaching question is only related to teachers that want to take the AAA and distribute lecture notes (or something) online. Even then, we cannot police local distribution. If they want to share it in a public forum (maybe a note website of sorts), then the modification would need to be CC. Global or local license:I think the local license is harder to maintain and it means that if we print a book, every single chapter will need a separate license to be printed. That's not that big of a deal, honestly. I think it's a good idea to attribute the algorithm-archivists and invite anyone who has a chapter submitted to the organization, itself. I don't know if this means we need an actual company or if a github organization would do. A global author file would be easy enough to create a license for:
Alternatively, if we claim the algorithm-archivists organization, it's:
For individual chapters, this might look like:
Again, that's fine to put on each webpage and isn't too annoying to print. We just need to remember to update the license for each chapter. Minor changes:I feel like people who make minor changes should be acknowledged, but we can do it like we said above:
Or even:
The problem is that if we update the license in the latter way (as more people come in), I am afraid we will need to get the permission of the original author, and that might be hard to do in some cases. By saying we have a global list of authors / reviewers, it is clear from the license initially that the initial author will allow modification from anyone on the changing list of authors. Contribution process:Clear guidelines are necessary, but having a in-build PR template for code submissions and chapter submissions with a box that says, "I accept that this contribution will be under this license." is a good idea going forward. |
Share-Alike / Non-CommercialI very much understand the thought process of the FSF behind SA and NC, but that doesn't really matter for this discussion. I would love to see the AAA being fully free, but I know about the risk, that not-NC (and not-NC in combination with not-SA), introduces. Global or local licenseThe algorithm-archivists organisation doesn't necessarily consists exclusively of all authors. If you want a global license a file with the contributors seems more reasonable. If we want to print the book, local licenses aren't really a problem. I think a smart solution can be figured out. For more info, look here under "Offline Work?". This page also suggests, that we don't need to include the CC license we're using. Linking to it seems fine. |
I just want to point out that I do not have a problem with that. I think the NC license |
No, that's still not what it's about. When A shares a work with B, B is granted the right to use it however they please, including the ability to share it with C. If that happens, A wants to make sure that C also receives all these rights. That's what Share Alike and licenses like the GPL are about. They require the rights/freedoms to be passed down the chain. They revoke the right to revoke rights. Copyleft and Share Alike have their downsides - no question. But they do make sure that the recipients of your work receive the rights you want to grant, no matter how far down the chain they are.
Share Alike definitely comes into effect when you use part of a work in your own work.
Source: Creative Commons Wiki
If a teacher uses CC SA licensed text in lecture notes, then they're required to license the lecture notes as CC SA in whole. That would mean that students would in turn be allowed to share the lecture notes freely. If the teacher decides to include other pieces with licenses incompatible with CC SA, then that could pose a problem. Though I'm pretty sure there is a way to separate the CC SA parts from the non-compatible parts in a way that could make it work (i. e. print the CC SA parts separately and merely reference them from the non-compatible part). Edit: I should probably point out that while my stance towards Share Alike seems positive in this comment, I'm not necessarily saying that we should absolutely use it. I usually publish my code with MIT rather than GPL and I like the the notion of imposing as few restrictions as possible. I'm just trying to explain that SA is not just restrictive for the sake of being restrictive. |
I don't want get into the SA debate again. I stand by what I said before. I am sketched out by it, but if everyone else wants to use it, we can use it. I still think the NC works here, but if we do individual licenses with different authors on every chapter, it becomes difficult to create a physical copy eventually because when we do, we need to contact all the authors who might have moved on. This is why a global attribution to the algorithm-archivists makes sense. We can then funnel all the money to the organization, instead of individuals, and then pay the individuals for dev time. I think it's alright to give attribution to everyone in the organization (for writing, reviewing and facilitating code contributions), but leave the small edits in No matter the case, we need to make attribution clear in the chapter contribution guide. I have no idea how to deal with SA with non-SA. Honestly, I just want to do whatever causes us the least stress in the future. |
ShareAlikeIf ShareAlike prevents the AAA from being used in educational scenarios (being incompatible with text-snippets from other papers and so on), we shouldn't use it. One key-goal of the AAA is enhancing educational content. NonCommercialI also think NC makes sense. The risk of companies abusing the AAA is just too high and the AAA organisation is so small, that we can't prevent it. Global or Local LicenseAfter thinking about it, a global license seems nicer, since the AAA is licensed as a whole. Just relicensing on every text change doesn't seem to bad. (We just need to make sure to preserve NC somehow (,if we want to use it).) |
I think I have changed my mind here. The only way for us to have exclusive commercial rights is with NC, but it seems like the NC license does more than what the wording of the license implies (here's a graphic to explain what I mean) I want the AAA to be used by as many people as possible. I just wanted to make sure people didn't sell a copy of the AAA to make a profit off the community without our explicit permission. I figure that it is a lot of work to take the current AAA and sell it, so it's probably fine to make everything SA so we can sell our own version (and I guess other people can sell a version too, but I doubt other people will). In order to keep the NC, we need to either cede commercial rights to me for the time being or create a company to have the commercial rights (which would be hard for me to do right now). It doesn't make sense to have more than one entity have exclusive commercial rights over different parts of the book. |
I believe most people are now more or less okay with the idea of re-licensing all AAA content under CC BY-SA, which can be done because I have exclusive rights over all written content right now. The main reason I am no longer in favor of NC is that under NC, the AAA cannot be used in tuition-based courses... Which is a bit restrictive. After this change is made, we can begin attributing authors and editors at the bottom of each chapter. In principle, the CC BY-SA license will allow anyone to print and sell copies of the AAA, but we will assume that they either will not do this or that we can provide a copy for less because we are not intending to make a profit off AAA sales. I will make this change soon, but would like to hear any further discussion before then. |
I was going to wait for the merge of the PR templates in #569 |
The only reason this issue is not closed is because the PR templates are not working right now. I will close this and open a new issue for PR templates |
As of right now, I (James Schloss) have exclusive commercial rights over the text in the Algorithm Archive (AAA). This was done deliberatively at the start of the project because I thought I would be the only author; however, since we have chapter submissions, this will no longer be the case and we need more rigorous chapter contribution guidelines along while also providing a license that "makes sense for each chapter."
I will explain the problem in as much depth as I can, so please forgive the length of this post.
Long-term goal:
I would like to be able to create a Print-on-Demand version of the AAA. Users select the chapters they want published and the language, and we print them physical copies. We should also provide guides through different topics (Like a "Guide to Computational Physics" or something), which provide a pre-bundled selection of chapters and some auxiliary text exclusive to each guide.
This is the only way I can actually see the AAA being sold, but I put the Non-Commercial tag in the Creative-Commons license because I didn't want people printing unofficial AAA copies and selling them.
The main problem:
There has been some discussion about whether the Non-Commercial is actually necessary in this case because:
In addition, we have one license over the entire AAA text (similar to how we have one license over all the code). If someone submits a new chapter to the AAA, they should be attributed for their work on the chapter. If someone edits the chapter, they should also be attributed. This is not currently how our CC license works.
Possible solutions to Non-Commercial problem:
No matter the case, the Non-Commercial tag seems unnecessary in this case; however, the discussion also came up about using Share-Alike instead. Share-Alike just means that any forks / modifications of the AAA must use the same license. It's one of those "more free licenses that also enforces freedom," so it is somehow more restrictive in another sense. I am 50:50 on the Share-Alike right now.
Possible solutions to authorship problem:
The authorship problem is a bit harder to tackle. We could do a number of things:
James Schloss et. al.
and define theet. al.
to be in a separate file calledAUTHORS.md
, similar toCONTRIBUTORS.md
for the code. If someone modifies the text, they can just put their name there. This has the advantage of only being in one place, so we don't need to put the license in every chapter online (and in print eventually).There are probably other solutions here that we discussed, but I missed. Please let me know and I'll update this.
Conclusion:
My current opinion is that we should remove the Non-Commercial tag in the license and share the license between everyone in a global
AUTHORS.md
file. If we want to sell bundled versions of the AAA later, this is perfectly fine without the Non-Commercial tag. As I mentioned, I am still 50:50 on the Share-Alike, but if you guys want it, I don't mind putting it in there.Please let me know what you guys think about this. It's kinda messy, but we need to get it right before merging chapters into the AAA.
The text was updated successfully, but these errors were encountered: