True, maybe I trusted the "latest" in the url too much ;-) In src/setup.py we set Cython.Compiler.Options.old_style_globals = True, which causes Cython to fall back to the old behavior.
possibly related: http://trac.sagemath.org/ticket/12446 I think the right place for the global variable injection would be the frontend-specific rich representation backends that I recently added in sage.repl.rich_output. Its dependent on the repl implementation after all. Then we could also do away with the Cython hack. On Wednesday, March 25, 2015 at 11:02:51 PM UTC+1, Jeroen Demeyer wrote: > > On 2015-03-25 19:40, Volker Braun wrote: > > Cython doesn't really have globals > > dictionary. > http://cython.readthedocs.org/en/latest/src/userguide/limitations.html > > says: > > > > "The globals() builtin returns the last Python callers globals, not the > > current function’s locals. This behavior should not be relied upon, as > > it will probably change in the future." > > This is about Cython 0.15 and the analogous document for Cython 0.22 > doesn't have anything anymore about globals(). Which makes me guess that > the fact that var() works is a bug and not a feature. > > -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
