Sorry if you don't want this kind of discussion here, and feel free to close without further comment if you don't.
Although it would ultimately be up for a court decide, there are many people who interpret the terms of the GPL in such a way that importing an R package in one's code creates "derivative work" of that package. This would mean that the importing code would also need to be placed under GPL if distributed.
I assume that many people who place their code under GPL want to prevent others from improving and then selling their code, without having to give back the improvements to the original project. They probably have much less of an issue with others creating new code that merely uses their libraries and then distributing that new code under whatever other license. It would not surprise me if many people who start out writing an R package actually think that the GPL actually allows this. See for example this occasion, where the authors of data.table seem to have thought this way and ended up changing their license after they became aware of this.
Please consider making a license specifically designed for libraries, such as the LGPL or a similar license (e.g. data.table ended up choosing MPL which is slightly different) the default when creating packages. New users who are likely to use the default option, and I would say the GPL is not a good choice.
Sorry if you don't want this kind of discussion here, and feel free to close without further comment if you don't.
Although it would ultimately be up for a court decide, there are many people who interpret the terms of the GPL in such a way that importing an R package in one's code creates "derivative work" of that package. This would mean that the importing code would also need to be placed under GPL if distributed.
I assume that many people who place their code under GPL want to prevent others from improving and then selling their code, without having to give back the improvements to the original project. They probably have much less of an issue with others creating new code that merely uses their libraries and then distributing that new code under whatever other license. It would not surprise me if many people who start out writing an R package actually think that the GPL actually allows this. See for example this occasion, where the authors of
data.table
seem to have thought this way and ended up changing their license after they became aware of this.Please consider making a license specifically designed for libraries, such as the LGPL or a similar license (e.g. data.table ended up choosing MPL which is slightly different) the default when creating packages. New users who are likely to use the default option, and I would say the GPL is not a good choice.