The 🅭🄍 Loophole That’s Been Annoying Me for 2 Years
Two years ago, I identified a loophole that involved the CC0 1.0 Universal Public Domain Dedication. I use 🅭🄍 for a lot of my work, and working around this loophole has been quite annoying. Luckily, the loophole has been fixed! In the U.S., written works are automatically copyrighted the moment they’re written. This can have some unintuitive consequences. For example, if I send you an email, you aren’t allowed to forward that email to someone else unless I give you permission (or you use an exception to copyright law). One way for me to give you permission is to write a document that says that you have permission. But wait. What if you wanted to forward that permission document to someone else? Well, the permission document automatically became copyrighted, so you need to get my permission in order to forward it. I could give you a second permission document for the first permission document! But wait. What if you wanted to forward the second permission document to someone else? Well, the second permission document automatically became copyrighted, so so you need to get my permission in order to forward it. I could give you a third permission document for the second permission document! But wait. What if you wanted to forward the third permission document to someone else? Well, the third permission document automatically became copyrighted, so… This recursive logic loop is especially a problem for FOSS repos. In order for a piece of software to be free or open-source, there needs to be some sort of document that gives certain permissions. Projects don’t typically write their own permission documents. Instead, they usually use a license that was written by someone else. Additionally, projects often include a copy of their license in their repo. Because license documents are automatically copyrighted, FOSS projects need to get permission in order to include them in their repos. Many licenses nip the problem in the bud. Some licenses explicitly give you permission to redistribute them: Other licenses implicitly give you permission to redistribute them: When I first discovered the loophole back in 2021, 🅭🄍 didn’t contain anything that would give the reader permission to redistribute 🅭🄍. Instead, there was a separate Web page that said: Now, there are two possible ways to interpret that quote. I interpret that quote like this: I know of at least one other person who interprets that quote like this: Personally, I think that my interpretation is more consistent with the rest of the text and is how the author interprets the text. That being said, I can’t prove that my interpretation is the correct one, so it’s possible that there was never any loophole at all. The problem here is that quote only talks about the legal code of 🅭🄍. At the time, Creative Commons’s Web site had a definition for “legal code”, and that definition excluded parts of 🅭🄍. That’s what the loophole was. Creative Common’s was probably trying to put all of 🅭🄍’s text into the public domain, but they ended up only putting part of it into the public domain. When I first discovered the loophole, the bottom of each page on Creative Commons’s Web site said: Based on that statement, I assumed that part of 🅭🄍 was in the public domain and that the rest of 🅭🄍 was available under 🅭🅯4.0. As a result, I was being diligent and including proper attribution, links to the author et cetera for the parts of 🅭🄍 that were copyrighted. It was really annoying because it meant either not including a copy of 🅭🄍 in repositories that use it or including a bunch of extra text alongside it. Now that I’ve found a workaround, it’s time to get the issue fixed upstream. Creative Commons has an official process for contributing to their projects. Here’s the gist of it: I opened two issues. And I waited, and waited, and waited. The first issue was assigned to someone pretty quickly. A year later, that issue was reassigned to two other people. Today, the issue is still open, and there’s no public signs that any work has been done on it. Additionally, that issue has been reserved for Creative Commons staff only. The second and more important issue still hasn’t even been triaged. While I was quietly waiting for these issues to be fixed, I found out about Jason Self’s WPDD. The WPDD is a fork of 🅭🄍, so it would also be affected by this loophole. I emailed Jason Self and tried to explain the loophole and how it could be worked around. Unfortunately, the conversation quickly devolved into an unproductive argument. I definitely failed to communicate effectively, and I never got permission to change Jason’s mind which was a huge mistake on my part. If you don’t get permission to change someone’s mind, then you’ll never be successful. There was one upside to that argument, though. That argument allowed me to reflect on how I was interpreting Creative Commons’s legal policies. I got a better understanding of what I think those legal policies mean and why I interpret them the way that I do. While I was having the argument, I realized that I might be misinterpreting those policies. In other words, I realized that my workaround might not work at all. I commented on one of the GitHub issues, and to my surprise, I actually got a response. The problem had almost been solved! Creative Commons had updated that policy to say: The only problem with this new statement is that it says “license pages”. 🅭🄍 is not a license. All they have to do is change that quote to say “legal tools pages”, and the loophole would be completely closed for good! Unfortunately, they haven’t gotten around to doing that yet, so we need another solution. One possible indirect solution would be to rely on the response that I had mentioned earlier. Part of that response says: All of the text that appears on the pages containing CC licenses and legal tools should be released to the public domain under CC0. (The licensing statement was written prior to clarifying which elements on those pages are and are not legal code!) I have two concerns with doing that: Another possible solution would be to find another place where Creative Commons talks about the legal status of 🅭🄍. If I can find somewhere else where Creative Commons clearly says that 🅭🄍 is in the public domain, then I wouldn’t need to wait for either of those issues to be fixed. I wasn’t able to find a direct reference to 🅭🄍 being in the public domain, but I was able to find this: That statement had two problems that were easy to fix. It also had two problems that were difficult to fix. For one thing, this statement has the same problem that the legal policy had. It talks about Creative Commons’s licenses, but it never says anything about 🅭🄍. I had first discovered this statement long before I had heard anything from Creative Commons about potentially changing their legal policy. I ended up waiting until after I got that response that I had mentioned earlier. Even once I had the response, there was still another problem that had to be fixed. At the time, that repo also said: So, I waited for longer. I forgot about it for a while, but recently, that warning was removed. I opened another issue, and to my surprise, it was fixed in a matter of weeks. As a result, there is now, finally after all of this time, an official statement from Creative Commons the clearly indicates that 🅭🄍 can be redistributed:What was the loophole?
How common FOSS licenses deal with this problem
How 🅭🄍 tried to deal with this problem
Tangent about the interpretation of that quote
The annoying workaround
The convoluted series of events that inadvertently closed the loophole
The attempted direct solution
The successful indirect solution
Finally, after all this time, I can remove all of the pointless attributions and explanations from my repos!