💾 Archived View for soviet.circumlunar.space › pjvm › log › 000011-creative-commons-nc-nd.gmi captured on 2024-05-12 at 15:38:37. Gemini links have been rewritten to link to archived content
⬅️ Previous capture (2024-02-05)
-=-=-=-=-=-=-
by pjvm
05-10-2023
commentary entry #2 (#8 overall)
Solderpunk announced that the license for the gemini and gemtext specifications will now be Creative Commons BY-NC-ND:
Separated specifications and copyright updates
I want to voice my disappointment with this, and explain why using such a license should be discouraged.
People often think it's a good idea to publish a work made in a hobbyist context under a CC "non-commercial" (NC) license. People also often think it's a good idea to publish official documentation and statements of opinion under a CC "no derivatives" (ND) license. Both of these are generally misguided.
Let's get one thing out of the way first: putting something under a license doesn't prevent people from doing what the license forbids. It just means that if they don't follow the license, you can try suing them for copyright infringement. The terms of a copyright license are worthless if you are not willing to take people to court over them.<1>
Now let's address the idea that ND protects the integrity of documents. Taking the gemini specification: you know the protocol itself is not copyrighted, right? What is copyrighted is the prose of the specification document, the particular wording used to describe the protocol. If you independently write a specification for gemini (or rewrite the official spec sufficiently to not be considered infringing in the applicable jurisdiction), that specification does not violate the copyright of the official specification. Another example: Wikipedia is able to give descriptions of closed standards without violating copyright. You can also write an incorrect specification that misrepresents the protocol, and there's nothing an ND clause will do to stop that.<2> Likewise, you can misrepresent people's opinions just fine without copyright infringement.
By the way, a quick note about relicensing, which Solderpunk already acknowledges: any contributions made under CC0 have been placed irrevocably in the public domain - you can license the document, but what that means is that the license becomes an alternative legal way to use the content, with the old legal route still available. So, to make your misleading specification, you can just use Sean Conner's protocol spec, insert your favourite protocol changes, and publish that - no need to write your own spec from scratch.
One more thing about ND is that it could be considered "open-washing".<3> ND makes a license so useless that it's not really preferable to putting it online as "all rights reserved"; doing the latter is then more honest.
In summary: copyright is not magic, and it is not a suitable tool to prevent misrepresentation.
Now on to NC. A common misconception of the NC clause is perpetuated by the author of xkcd:
If you're not sure whether your use is noncommercial, feel free to email me and ask
This is incorrect: the definition of "noncommercial" is not determined by the author of the work. What actually matters is how a court will interpret and apply the license's definition for the term.
Of course, Munroe can absolutely give you permission to use xkcd in some way when the license does not allow this use; direct permission from the copyright holder means you do not need the license. What a copyright holder cannot do is declare some use commercial when the license doesn't define it as such.
Now let's look at the definition given in the license:
NonCommercial means not primarily intended for or directed towards commercial advantage or monetary compensation.
This seems like a perfectly reasonable definition of the term "noncommercial". The problem, as I see it, is that no definition can ever be good enough.
First of all, let's consider what this disallows that you may want to allow. As an example, because of the "monetary compensation" bit, one thing that people will typically think they are not allowed to do is to sell physical copies at cost (i.e. without making any profit). Now, whether a court would actually rule that the license doesn't allow that, is another matter - as long as the license can be interpreted as forbidding a certain use, some people will interpret it that way. Even worse, people will not always actually go and read the license; instead, they may intuitively conclude that their use is "commercial" by their own definition and therefore not allowed.
Now of course, people can always ask you directly if you approve of their use - but how many actually will? You've chosen that specific license for a reason, so there is no point in asking; or so people will think. Ironically, when a work is not licensed, people might be more aware that they can ask the copyright holder for permission. Also, presumably the reason you would put a license on the thing is so people don't have to go directly to you all the time.
Next, let's consider what you might want to disallow that the license might actually allow. The "not primarily intended" part of the definition seems particularly dangerous here: for how many uses might it be possible to argue that making money is only a secondary purpose, and the primary purpose is something else? Also, if I understand correctly, the part "intended for or directed towards" might be hard to prove and (like anything) could be jurisdiction-dependent.
You can not disallow a use if it falls under fair-use exceptions (which are jurisdiction-dependent), if you are not willing to sue someone over it, or of course if you never find out that it happened. Another problem is that people might naively violate the license, reasoning intuitively that their use is noncommercial while the license (as interpreted by a court) disagrees.
The usual motivation for choosing a NC license is: you have some vague idea that if you don't, a for-profit corporation might take your stuff and make money off of it somehow, and that idea feels icky. The actual likelihood of this scenario typically varies between small and non-existent.
Meanwhile, you can achieve something basically as good by just using a "share-alike" clause (SA; more or less the equivalent of copyleft). This makes it so any derivative of your work will stay freely distributable. It's not an original observation that share-alike achieves more or less the same as noncommercial, with fewer unintended consequences.<4> The CC BY-SA license already makes it near-impossible to make a profit from parasytic behaviour like asking money for digital distribution, because people can always get the same thing for free. Even asking money for physical copies becomes pretty hard, unless you do it at cost or do something like "90% of profit is given to the author".
In summary, if you are thinking of using a NC license, you can likely better achieve what you want by doing one of two different things: either putting it as "all rights reserved" with an indication that people can always ask permission for things and some examples of what you would give permission for; or just using BY-SA.
In conclusion, I think putting the gemini and gemtext specifications under CC BY-NC-ND is misguided and pointless; nothing of value is achieved by this. What it does achieve is that fewer people will be willing to make contributions to the gemini project. Myself, for example: in the past, I did some translation work of gemini documentation into Dutch, but in the future I will not.
<1> I'm somewhat tempted to put up an unofficial spec translation up on my capsule. Would Solderpunk sue me (or the sysadmin of this domain) over that?
<2> Again, somewhat tempted.
<3> I'm of the opinion that it was a grave error for Creative Commons to create the NC and ND licenses; it gives legitimacy to such licenses and allows people to say they "use a Creative Commons license", which is utterly meaningless but sounds benevolent.
<4> I specifically recall reading this in a document produced by Creative Commons itself.
PS: The title of this post is a quote from a social media post that I remember reading but cannot find back. If you happen to have the link to said social media post, I would appreciate it if you sent it to me so I can properly reference it here.
Disclaimer: I am not a lawyer and you shouldn't use this post as legal advice
Addendum 05-10-2023. clseibold has posted a reply:
gemini://bbs.geminispace.org/u/clseibold/5895
I'll respond to a few parts.
Solderpunk made his intentions clear, that he didn't want derivative specs that *look like* the official spec spreading around that made changes and fooled people into believing it's an official spec
They seem to have missed the parts where I explained you can still do exactly that? I talked about how you can make a document that looks like it's an official specification and put in misleading details; in the case of gemini it's even easy, because you can use the public-domain protocol spec as a base instead of writing a whole independent spec. A bad actor is not restricted to publishing summaries, and copying the wording of a document is not necessary to create something that can masquerade as that document.
If the copyright holder is more lenient than the license, then there's literally no problem
Well, this depends: has the copyright holder thought about what they want to allow vs what the license allows, and publicly stated that certain things are allowed? If not, yes there is a problem: people will assume what the license allows is exactly what the copyright owner intends to allow. clseibold is presuming here that the copyright holder has been careful with the license choice and fully understands what the consequences of the license are - but that assumption often does not hold true.
ND's use is when one wants to disallow derivatives while still *allowing unmodified redistributions*.
Yes, I know. However, I think that's a pretty useless permission to give when the work in question is publicly available on the internet (with one exception, see below). With all rights reserved you can also link to it, crucially you can also archive it, and you can distribute copies of it in private communication.
ND and NC do actually have good uses, particularly outside of text, but this whole post just judges these licenses based only on the idea that the only artform that ever exists is writing
While I've used writing as my examples here, that doesn't mean I was assuming writing is all that exists.
One thing I had not considered is that ND allows the work to appear unmodified as part of a larger whole. For example, putting a comic as an inline image in an html document. So I guess they have a point: for images, it could genuinely be that ND is not a mistake on the licensor's part, and that the license does allow something useful. Nevertheless, I would not call the licensor's intention a "good use" there.
Another thing not considered in my post is that NC can be used to allow the copyright holder to make money. Here, an example for visual arts would be to reserve the exclusive right to sell merchandise. I'm not sure I would consider this a "good use", but I can see the thought process. However, such a use of NC is then in a professional context, whereas my post is really about naive use of licenses in a hobbyist context.
A license is still useful by telling people what your wishes are, even if you do not have the means or desire to sue people.
If you only want to tell people your wishes, you can simply state your wishes. A copyright license is a legal instrument.
Lastly: even if clseibold were right and putting the specs under CC BY-NC-ND did achieve something, you also have to consider what might be lost by doing so, and evaluate the trade-off. And even if everything they said were correct, I still think the trade-off wouldn't be worth it.
Edit 05-10-2023. A previous version of my reply to clseibold did not consider the "comic in a blogpost" example and the "selling merch" example.