>>>>> On Tue, 10 Mar 2026, Michael Orlitzky wrote: > On 2026-03-10 18:58:36, Ulrich Müller wrote: >> Not sure if LICENSE would be the right tool for this. AI generated code >> certainly touches legal aspects, but I think these are not at the core >> of the issue.
> I was going to suggest a LICENSE-based approach as well. Not everyone > will want to avoid these packages for the same reasons, but tagging > them with (say) LICENSE="dubious MIT" would be semantically accurate, > and would not hinder further refinement. > If the upstream repo might contain an algorithm plagiarised from a > copy of TAoCP on libgen, then the license isn't "MIT" no matter what > the Github sidebar says. What is it? Well... we don't know yet. It > will probably take a few decades to iron out. In the meantime, it's > dubious. IMHO there's nothing dubious here. An algorithm by itself (i.e. its ideas or conceptual steps) is not copyrightable, so you can take any algorithm from TAoCP, as long as the implementation is your own. On the other hand, a concrete implementation of the algorithm is copyrightable, so a verbatim copy from the book would not be allowed. In my (paper) copy of TAoCP I find this: | Copyright © 1998 by Addison-Wesley | All rights reserved. No part of this publication may be reproduced, | stored in a retrieval system, or transmitted [...] without the prior | consent of the publisher. So if a package would contain any significant amount of material from TAoCP then it seems pretty clear that we needed "all-rights-reserved" in its LICENSE. Ulrich
signature.asc
Description: PGP signature
