[int_19h]

For Microsoft GitHub repos, there's some kind of automated tool that enforces CLAs. Whenever there's a new pull request, it checks whether the author is a Microsoft employee or not. If not, it checks whether there's a CLA for that person, and if there isn't one, posts a link to a page that handles signing, and records the result for future uses. Repo owners are also notified accordingly.

It also seems to have some kind of threshold for code size, because for small contributions, it says that CLA was not required.

You can see it in action from users' perspective in pretty much any repo under https://github.com/Microsoft/, if you search for "MSBOT" in pull requests. E.g.:

https://github.com/Microsoft/PTVS/pulls?page=1&q=is%3Apr+MSB...

(note the labels)

[rando832]

And note the Microsoft CLA says "you grant Microsoft a patent license covering your contribution", while Microsoft grants no patent licenses over (edit: most) of their open source code, and are currently profiting from patent licenses against open source code in android etc.

When it comes to patents, MIT is a closed source proprietary license, while Apache 2.0 and GPL are open source and free software. All these companies spewing bs about how open MIT is bugs me to no end.

[daveguy]

It takes a really warped interpretation of open source to claim MIT is a closed source license. Not even Stallman agrees with that. In his essay on selling exceptions to GPL he gave this argument[0]:

" If that implication is valid, it would also apply to releasing the same program under a noncopyleft free software license, such as the X11 license. That also permits such embedding. So either we have to conclude that it's wrong to release anything under the X11 license -- a conclusion I find unacceptably extreme -- or reject this implication. Using a noncopyleft license is weak, and usually an inferior choice, but it's not wrong.

In other words, selling exceptions permits some embedding in proprietary software, and the X11 license permits even more embedding. If this doesn't make the X11 license unacceptable, it doesn't make selling exceptions unacceptable "

It seems everyone except you, including the father of copyleft, agrees that the MIT license is a free and open source license.

[0]:https://www.fsf.org/blogs/rms/selling-exceptions

[pjmlp]

Other than I can take a MIT licensed project used to build my product, sell it and never give the anything back, not even the code.

Seems pretty close source friendly to me.

[opk]

And many open source developers are just fine with that. If someone finds a use for my code and even if they make lots of money off it then good for them. It's not like they've stopped me from using it in the original form. I've also had cases at work where we want to use open source and copyleft licences have caused clueless managers to get panicky despite the fact that we don't redistribute anything, only make use of it internally. They also often initially tell me not to contribute anything back but after the work is done and I explain multiple times about the effort of porting the changes to subsequent newer releases they agree (often with the proviso of using my personal e-mail). In summary, I don't follow the RMS line of closed source being morally evil or whatever so a license like ISC or MIT is the best practical match to my wishes when I release stuff.

[int_19h]

>> while Microsoft grants 0 patent licenses over any of their open source code

PTVS is licensed under Apache 2.0.

https://github.com/Microsoft/PTVS/blob/master/LICENSE

[rando832]

I stand corrected.

[johncolanduoni]

Most of the big MIT licensed Microsoft projects I know (.Net Core VM and libraries) are licensed under MIT with a separate, explicit patent grant. For example look at the PATENTS.TXT file in this repo: https://github.com/dotnet/corefx

[rando832]

Interesting. It's a non-legally binding patent promise. But honestly, I think MS has waded too far into open source to have the bad publicity of breaking that promise, unless MS gets into very bad finances. Anyways, good enough for me.

And it actually makes my point about the MIT license, it's inadequate for patents.

[johncolanduoni]

What makes you think it's not legally binding? As far as I can tell it's as binding as the MIT provisions.

[rando832]

Looks like I was wrong.

According to these links, they will probably work in an actual case. They can be revoked at any time, making them inapplicable to future patent use which was not already going on for some period of time.

http://en.swpat.org/wiki/Equitable_defence, http://en.swpat.org/wiki/Patent_promises

Definitely not anywhere near as binding as the MIT license, but it's not a good comparison anyways, MIT addresses only covers only copyright and patent law is a very different and separate area of law.

[rando832]

As others have pointed out, MS does have explicit patent promises, so my characterization was wrong.

[vsl]

> It also seems to have some kind of threshold for code size, because for small contributions, it says that CLA was not required.

That's because all this copyright stuff is needed only for things that can be protected by copyright in the first place. One of the elements required is creativity of the work and if you only make a trivial change, e.g. Adding an obvious make file rule or adding a file to a project, that's not creative (there's just one obvious way to do it), so not protected, so not needing a CLA.