GitHub's licensing data illustrate the decline of copyright monopoly

[Arto Bendiken‏]

Good morning peeps,

GitHub have yesterday published some pretty detailed licensing
statistics [1] for the circa 20.4M open-source repositories [2] hosted
on their service. The growth in Unlicense adoption is clearly
discernible: we're at soon 2% of "licensed" projects (or, about 1 in
50 repositories from this sample). That's a total of 70,000+
public-domain projects, more every single day [3].

I'm waiting to get the exact CC0 figures, but indications are [4] that
the combined Unlicense and CC0 figure is about 2.5%, or 1 in 40
"licensed" repositories (the qualifier is important, and will be
explained below). That's pretty good compared to only a few years ago
when the public domain simply wasn't on the map at all.

To put this in perspective, the Unlicense and CC0 combined are already
at 40% of the combined BSD license (2-clause/3-clause) figures. From
yet another perspective, Unlicense adoption by GitHub users is already
larger than that of LGPLv3 and already more than a fifth (⅕) that of
GPLv3 adoption.

As a return-on-investment metric, that's rather rewarding to those of
us who founded the Unlicense initiative. After all, Mr. Stallman has a
budget and millions of rabid true believers, while we can hardly be
bothered to tweet, much less make more noise than that. We have been
relying on both apathy and word of mouth to spread an idea whose time
has almost come, and this is indeed bearing fruit. (I wonder what we
could do if we actually put some effort into it.)

Still more noteworthy is that more than 80% of all GitHub repositories
don't have any license at all, and the supermajority of the ~20%
licensed repositories are using a highly permissive license (chiefly,
the MIT license). Copyleft is withering away, as is the copyright
monopoly as a belief system. Good riddance.

GitHub's Mr. Balter, who spearheaded the initiative to collect these
stats, thinks it unfortunate [1] that people are increasingly choosing
to not give a shit about licensing, and encourages everyone to add a
licensing statement to their repository. That's well and proper per
his job description of looking out for his employer's and customers'
interests as he perceives them. However, as the data he's collected
clearly show, you just can't hold back the tide—we're rushing headlong
into a post-copyright future, and we're past the point of no return.

Given that the Unlicense was always intended to be merely a transitory
instrument, a way of backporting the copyright-free future into an
easy-to-use format suited to the copyright-afflicted present [5], the
80%+ figure excites me still more than the success the Unlicense has
enjoyed thus far. I look forward to the eventual day when we can
simply retire the initiative, and it's getting closer every year.

In the meanwhile, there are a number of tweets at @TheUnlicense [6]
that you may interested in examining and retweeting.

Thanks for your attention,
Arto

[1] https://github.com/blog/1964-open-source-license-usage-on-github-com
[2] https://github.com/about/press
[3] https://github.com/search?o=desc&q=unlicense&ref=searchresults&s=indexed&type=Code
[4] https://twitter.com/theunlicense/status/575068146400391168
[5] http://ar.to/2010/12/licensing-and-unlicensing
[6] https://twitter.com/theunlicense

--
Arto Bendiken | @bendiken | http://ar.to

[Peter Saint-Andre‏]

Hey Arto, that's great stuff and some interesting observations.

Peter

[Tomas Pluskal‏]

Hi Arto & everyone,

While I generally agree with the opinion that the copyright concept is outdated, I would also like to express some scepticism regarding the application of Unlicense. I see that most GitHub projects are "free" and "for fun" type of projects whose authors generally don't care much about legal details (hence 80% of the repos exist in legal void by having no license at all, nevertheless people happily share and exchange their code anyway). I also participate in various software projects that often combine libraries and components published under incompatible licenses (e.g., GPL + Apache + BSD etc), which is obviously illegal, but nobody really cares.. So, in my opinion, for this kind of personal projects using Unlicense is pretty much equivalent to using no license at all, because, as I said, nobody really cares.

But there is also the other world, the world of commercial software. For obvious reasons, commercial companies must be very strict about obeying laws, including copyright laws - no matter whether they like the law or not. Even if the company is willing to publish their software for public use, they cannot risk future legal actions against themselves. To give an example: a company based in the U.S. publishes an open-source software under Unlicense. Later, a volunteer programmer from Germany contributes new code to the project. The company grows bigger and distributes many copies of the software. Now much later in the future, if the German programmer suddenly decides to demand royalties from the company, does Unlicense provide any protection for the company? I doubt it, and (as far as I know) it has never been tested in court.

The Unlicense was crafted by non-lawyers and from its website it is not clear whether it was even reviewed by a lawyer or a specialist on international copyright law. It seems to be particularly problematic in jurisdictions where the concept of public domain does not exist or it is not legally possible to dedicate something into public domain (I mentioned Germany as their law seems to be the most restrictive).

The copyright law develops much slower than the technology, so we can only hope it will catch up eventually (and globally). But in the meantime, I find it hard to see who can really benefit from using Unlicense. Small-scale, non-commercial programmers can simply use no license at all and the effect for them is pretty much the same. But companies that require certain level of legal security cannot get it from Unlicense.

[Note: this is not an attack on the mission of Unlicense. I am simply concerned about the actual consequences of using it, when it comes to legal actions.]

Best regards,

Tomas

[jean-marc lienher‏]

Hi,

Tomas Pluskal a écrit :
...

> application of Unlicense. I see that most GitHub projects are "free" and
> "for fun" type of projects whose authors generally don't care much about
> legal details (hence 80% of the repos exist in legal void by having no
> license at all, nevertheless people happily share and exchange their
> code anyway).

They don't care about legal details now. But what will they say if
somebody makes money with their projects? They will certainly request
royalties.

These projects are not in legal void, they are automatically copyrighted
to their authors. Since their authors don't explicitly allow
redistribution of these works, nobody can redistribute them.

> I also participate in various software projects that often
> combine libraries and components published under incompatible licenses
> (e.g., GPL + Apache + BSD etc), which is obviously illegal, but nobody
> really cares.. So, in my opinion, for this kind of personal projects
> using Unlicense is pretty much equivalent to using no license at all,
> because, as I said, nobody really cares.

Somebodies do not care. But there is a lot of people who cares.
If a project has no license then I ask the author under which terms it
is distributed.
I would never aggregate a unknown licensed project to my projects.

>
> But there is also the other world, the world of commercial software. For
> obvious reasons, commercial companies must be very strict about obeying
> laws, including copyright laws - no matter whether they like the law or
> not. Even if the company is willing to publish their software for public
> use, they cannot risk future legal actions against themselves. To give
> an example: a company based in the U.S. publishes an open-source
> software under Unlicense. Later, a volunteer programmer from Germany
> contributes new code to the project. The company grows bigger and
> distributes many copies of the software. Now much later in the future,
> if the German programmer suddenly decides to demand royalties from the
> company, does Unlicense provide any protection for the company? I doubt
> it, and (as far as I know) it has never been tested in court.

The Unlicense is based on the SQLite public domain dedication. It seems
that big U.S. companies don't have any problem in using SQLite in their
products.

I think that you can never have worldwide protection with a unique
license. The laws are different for each countries, you must adapt your
license for each jurisdiction.

> The Unlicense was crafted by non-lawyers and from its website it is not
> clear whether it was even reviewed by a lawyer or a specialist on
> international copyright law. It seems to be particularly problematic in
> jurisdictions where the concept of public domain does not exist or it is
> not legally possible to dedicate something into public domain (I
> mentioned Germany as their law seems to be the most restrictive).

I think that a clever court in any jurisdiction will understand the
concept of public domain. "do whatever you want with it & no guarantee
at all".

As the author of illegal source code you will always be liable under law
for your acts, whatever the license under you distribute your code.

>
> The copyright law develops much slower than the technology, so we can
> only hope it will catch up eventually (and globally). But in the
> meantime, I find it hard to see who can really benefit from using
> Unlicense. Small-scale, non-commercial programmers can simply use no
> license at all and the effect for them is pretty much the same. But
> companies that require certain level of legal security cannot get it
> from Unlicense.

I disagree. No license equals "redistribution not allowed".
Public domain or Unlicense equals "do whatever you want".

I don't think that the Unlicense is worst than BSD, MIT, BSL...
If you need absolute level of legal security then you must write all
your code by yourself.

Best regards,
Jean-Marc Lienher

--
http://cod5.org