I’m a former user of root and I was using it quite extensively throughout my time as a PHD student at CERN.
I have a python re-implementation in place of a small part of root, which I’d like to distribute as a standalone python package. Can you please advise me which license I have to use in such a case and in what extend I have to link to the original root-code?
Good to hear that ROOT was useful, thanks for sharing!
Most of ROOT’s own code (i.e. not contributed packages such as cppyy) are LGPL’ed. I’m no lawyer, and so I cannot answer your question. But I have a personal opinion / expectation, and that is that a re-implementation does not depend on the license of the original software, but that we would certainly appreciate a “kudos” in publications / documentation / code!
Out of curiosity, which part are we talking about?
from what I understand, I think I have to use at least LGPL2.1, or higher, right? And of course, documentation etc. would include kudos to the original code
The re-implementation is basically about the functionality of the TGraphAsymmErrors.Divide() function, so basically the calculation of uncertainties for efficiencies and 2 Poisson means.
ROOT is indeed licensed LGPL 2.1 or higher. My opinion / understanding on whether a reimplementation has to have the same license didn’t change, now that I know it’s just about a couple of additions, multiplications, and divisions!