Move from GPL-3 to MIT license

[etiennebacher]

I have a package that reuses the code for the NSE developed in datawizard. datawizard and all other easystats packages are licensed with GPL-3, which means that I also need to license my other package with GPL-3 instead of the more permissive MIT licence. Is there a reason to have GPL-3 licence for all easystats packages? If not, is it possible to switch them to MIT like the tidyverse did?

I didn't see any issues about licence in this repo, so I hope you didn't have this conversation before.

Progress tracker

• insight
• modelbased
• bayestestR
• effectsize
• parameters
• performance
• correlation (Change LICENSE to MIT correlation#300)
• report
• see (Relicensing {see} see#305)
• datawizard (Switch from GPL-3 to MIT license datawizard#450)
• easystats (Change to MIT licence #383)

[IndrajeetPatil]

Yeah, I agree that we should switch to MIT.

We started with GPL-3 because that was the prevalent community norm when we started (tidyverse packages also included that license at that time). But I think we can relicense now.

Let others benefit from our codebase the same way we have benefitted form others'.

[etiennebacher]

I'm wondering how to do this process. For transitioning tidyverse packages, they collected the agreement of all contributors which had "significant" contributions (not typo fixes). For young packages like datawizard this might be easy to do but for larger/older ones it's more annoying.

Or do we switch just on our own?

[strengejacke]

datawizard and all other easystats packages are licensed with GPL-3, which means that I also need to license my other package with GPL-3 instead of the more permissive MIT license.

First, you can mix licenses, you don't need to change your package license.
Second, since it's "our" code you're re-using, no one will sue you, so just take the code.
Finally, I'd say if others made substantial contributions, like a complete function (see https://github.com/easystats/parameters/blob/main/R/dominance_analysis.R), we should ask. For existing code that has been slightly modified, we can decide on our own...

But: IANAL

[DominiqueMakowski]

Agree with all the above

[rempsyc]

Agree it's 2023 let's change to MIT

[bwiernik]

Agreed. Legally, we need to get consent from authors of copyrightable code (which is anything substantive, more than typos) to change the license, or reimplement their code if we can't get a hold of them or they don't agree.

[IndrajeetPatil]

If you are unsure about the steps needed to relicence, have a look at this: https://r-pkgs.org/license.html#relicensing

I think it shouldn't be too difficult to do this switch at the current stage.

[IndrajeetPatil]

bump

Can we please prioritize this for the next release?