ACK and go, when Travis agrees.

Uhm, are we allowing gplv2 into RIOT now?

Was there ever a reason against it?

I mean for a particular board support or driver it should be okay in a single file.

Uh, I'm with Ludwig here, for the code that ends up on the board we've been only allowing lgplv2 and compatible so far. For the tools it doesn't matter,

@LudwigKnuepfer

Uhm, are we allowing gplv2 into RIOT now?

@kaspar030

Uh, I'm with Ludwig here, for the code that ends up on the board we've been only allowing lgplv2 and compatible so far. For the tools it doesn't matter,

Well, with boards/x86-multiboot-common/include/multiboot.h there is already a GPLv2 file in here ;-)

Here is a newer release of multiboot.h with an MIT license.

Well, with boards/x86-multiboot-common/include/multiboot.h there is already a GPLv2 file in here ;-)

Not that the license would matter for a header. ;-)

\o/ then let's take out the GPLv2 pattern of this PR and change multiboot with the updated version in another PR.

@Kijewski are you free to do that?

@authmillenon, nope not right now, sorry. But ping me if some code needs to be adapted for the new file.

I have no capabilities right now either… :(

Slackers! ;-)

Updated the header.

How can I test this replacement works?

I don't know. I simple tried default example on i386-qemu, but the shell seems to be broken anyway.

Confirmed (might be because it does not implement uart_stdio?). ACK. Please squash

squashed

The purpose of the license checker is to warn about files with non-compatible licenses.
Of course we are at liberty to ignore this warning in certain situations, but in order to not let RIOT become riddled with non-compatible parts we need the checker to continue warning us.

Not that the license would matter for a header. ;-)

As far as my copyright knowledge goes (which is not much I have to say) the documentation needs also be freed from the authors copyright via the license. Otherwise, the sole copyright lies with the author and we wouldn't be able to take the documentation for our own. IANAL, but regarding this: yes, license in a header does matter.

The purpose of the license checker is to warn about files with non-compatible licenses.

But GPLv2 is compatible with LGPLv2.

This PR doesn't introduce GPLv2 checks so I merge now.

The purpose of the license checker is to warn about files with non-compatible licenses.

But GPLv2 is compatible with LGPLv2.

Have to be careful with those statements. Distributing binaries with GPLv2 inside requires the whole code to be distributed as GPLv2. In this case, GPLv2 is viral. I don't think that's what most people expect when talking about "compatibility" of licenses.

Distributing binaries with GPLv2 inside requires the whole code to be distributed as GPLv2.

That's why I wrote "I mean for a particular board support or driver it should be okay in a single file." Particular for the qemu (or native) support, I don't see reasons against GPL.

In this case, GPLv2 is viral.

Is there any case where GPL is not viral?

I don't think that's what most people expect when talking about "compatibility" of licenses.

The stupidity of the mass is not my fault. ;)