Ludovic Courtès writes: > Hello! > > Amin Bandali skribis: > >> On 2018-12-17 9:29 AM, Efraim Flashner wrote: >>> On Sun, Dec 16, 2018 at 04:18:16PM +0100, Ludovic Courtès wrote: > > [...] > >>>> > +(define-public python2-z3 >>>> > + (package (inherit python-z3) >>>> >>>> This definition cannot be in python.scm; it must be in the same file as >>>> ‘z3’ or we can get “unbound variable” errors while loading either of >>>> these two modules. >> >> Oh I see. If we choose to keep it (add it), I’ll move it to maths.scm. >> >>>> Also, as we’re approaching end-of-life upstream for Python 2.x, we now >>>> avoid creating “python2-” packages, unless we cannot avoid it for some >>>> reason. Do you think we could do without this “python2-z3” package? >>>> >>> >>> Currently our z3 package builds python2 bindings >> >> What Efraim said. Since the current z3 provides python2 bindings, I >> thought I would preserve that option by adding a python2-z3 in case >> anyone wants to continue to use the python2 bindings. >> >> I’m Cc’ing Marius who’s one of the recent committers to the z3 package >> definition. Marius, any thoughts on whether we should keep the python2 >> bindings around or do away with them? > > Marius, WDYT? Hello! I don't actually know z3 at all, I just updated it to fix the build on core-updates :-) In any case dropping python2 bindings seems sensible, seeing as Python 2 is EOL in a year[1]. So please go ahead, thank you Amin! [1]: https://www.python.org/dev/peps/pep-0373/