When one looks at GitHub, permissive licences like MIT and BSD-3 are favoured over the weak and strong copyleft licences from the GNU libraries (for instance, back in 2015, MIT-licensed repositories accounted for twice as many repositories as all GNU licences combined). Also, there seems to be a trend towards permissive licences against copyleft ones over the past few years.

Taking aside that the popularity of a licence encompasses many different aspects[Note 1] that go beyond the scope of what we can humanly assess here, what are some of the downsides to copyleft licences that would make a developer wary of choosing them for their software?

[Note 1] Two reasons I came up with that are likely to account for their popularity are the ease for a small developer to simply provide a short, simple and direct permissive licence like MIT instead of having to go through and learn all the lawful/verbose details on GNU licences, as well as corporations providing the backbone of their frameworks with permissive licences, but then likely developing their own closed-source versions built upon them.

Either way, what I am trying to understand here is not the incentives to choosing a permissive licence, but rather the disincentives to choosing a copyleft one.


Here's one specific example: in corporate environments where a company has in-house software that they want to keep proprietary, it would be a disaster if someone uses an external library in a way that legally requires the company to reveal the source code of its in-house software. When using an external library that has a copyleft license, if the library gets linked with the in-house software which is then distributed, that's exactly what would happen. While it is possible to avoid it by being careful not to link the wrong things together, it's much easier to avoid it if you don't use the copyleft software at all. That often provides a strong incentive to avoid copyleft software at a private company.

Permissive licenses like the MIT license do not carry that risk, so companies are generally much more willing to use software that is under a permissive license in their products.

As a result, if you're writing a piece of open-source software and you want it to be used as widely as possible (i.e. by private software companies, if the companies find it useful), then that goal is much better served by publishing the software under a permissive license than under a copyleft license.

  • It is interesting to note that while they are all licences for open source software, their "end goals" (in lack of a better term) differ greatly: permissive licences aim for maximum reproducibility of code, while copyleft aim for maximum openness of code. – mepqfilho Jan 10 at 12:53

In my point of view is all related to the reciprocal complexity that some copyleft licenses requires.

In the other hand, most of the GitHub projects are small projects that has low complexity and don't need copyleft licenses or the maintainer don't have the knowledge to decide/understand which copyleft license to use.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.