On Wed, May 6, 2009 at 4:29 AM, William Stein <wst...@gmail.com> wrote: > Suppose you spend three years implementing an algorithm as part of > Sage to compute X (say some Monsky-Washnitzer cohomology > computations). Then somebody else writes and publishes a clever paper > that includes a several-page Sage program that uses your > implementation of X (plus many other things in Sage) to compute Y (say > p-adic Regulators of Jacobians of genus 2 curves). Would you > definitely be allowed to use their new code and include it in Sage?
Suppose you spend three years writing a proof of X (say, some very important result about high rank elliptic curves), which you publish as a paper under cc-by-sa in some journal. Then somebody else writes and publishes a very clever paper that includes a several-page mathematical proof which uses your implementation of the proof of X (i.e. statement of lemmas, propositions, etc. leading to X which appear in your paper), to prove the BSD conjecture. Would you definitely be allowed to republish their new proof and include it in (say) wikipedia? If this were true, some obstinate mathematicians could be releasing their proofs under the GPL, thus eventually "infecting" all mathematics... (provided they reach critical mass, say). Wrt the "clever paper" you mention above, I can see two possibilities, maybe depending on the interpretation of the law (the notion of derivative work) by a court: (a) the paper is NOT legally a derivative work of Sage. Then the journal is clean, and you cannot claim anything about it. (b) the paper IS legally a derivative work of Sage. Then the journal is clean or not depending on whether they have a license to reuse Sage. Assume for the sake of the argument that the GPL is the only valid license Sage the editorial can obtain. The GPL is a permission (license) to reuse Sage which is valid as long as they comply with the terms. Then either (b1) the journal complies with the GPL by releasing the new code as GPL ===> you would be definitely allowed to use the new code and include it in Sage; or (b2) the journal doesn't release the new code as GPL ===> then the journal would be infringing Sage copyright. It does *not* follow that you are allowed to use the new code and include it in Sage. It *only* follows that the copyright owner for Sage can sue the editorial for copyright infringement, i.e. distributing a derivative work without a license. (IOW, the GPL is *not* really viral --- copyright law is). Remember: the GPL is only a *license*. Nobody is forced to agree, and nobody is forced to comply. The GPL gives permission to do some things that copyright law would not allow, provided its terms and conditions are satisfied. If I want to do something that copyright law alone would allow me to, then I don't need to follow the GPL. This includes fair use. For instance: what's the difference between referencing the sage library by using an import statement and a function call, vs. referencing a theorem in a paper by citing it in the bibliography and recalling the statement? They both seem fair use to me... In principle: I guess it would be possible to write a paper which cites, references, and recalls so much from a previous paper so that it becomes a derivative work and fair use doesn't apply; the same is true for scripts which reference the sage library. IOW: I'm claiming that *some* scripts can be derivative works, but others are either not derivatives or under fair use (this has been already pointed out by Robert Bradshaw). I'd say *most* notebooks making use of the sage library in the intended way, using the documented APIs, etc, would fall into the second category, and thus can be published under any license or no license at all. Gonzalo --~--~---------~--~----~------------~-------~--~----~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to sage-devel-unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://www.sagemath.org -~----------~----~----~----~------~----~------~--~---
