On Tue, Jan 14, 2020 at 9:42 PM Nils Bruin <nbr...@sfu.ca> wrote: > > On Tuesday, January 14, 2020 at 9:01:42 AM UTC-8, Frédéric Chapoton wrote: >> >> So here is my proposal. >> >> * Starting from now, we allow ourselves to move on, using 9.1 betas and >> further releases for external python3 updates, including switch to ipython7, >> which seems to me the most urgent matter. But we also do not introduce >> python3-only code in our own code base if we can avoid it. >> > > I think for development and stability, we should take heed of the fact that > trac has some recent tickets addressing issues with py3 that are only being > uncovered now. The reality is that prior to 9.0, sage/py3 got only very > limited testing outside doctests and developers working on it. Doctests by no > means cover the full functionality of sage, and I would expect that in the > next few months a lot more issues will be found. I'd be a little hesitant > dropping py2 support under those conditions: being able to drop back on py2 > in these situations for a while seems like a decent mitigation strategy, and > would probably help with debugging. > > I think similar thinking led to the original proposed and published strategy > on the wiki, and it still makes sense. In practice, sage/py3 is not really a > proven product yet; just because people won't start really using it until > there's actually an official release with it. > > I appreciate the good intentions of "not introducing py3-only code in our > own code-base if we can avoid it", but how much does that help? If there's > one such change in a file that gets loaded into sage upon start-up, or which > needs to be parsed during build, the possibility of running sage on py2 is > gone, isn't it? > > I agree it's perhaps frustrating that the process of moving sage from py2 to > py3 hasn't completed with py2 falling out of support, but having a first > official release based on py3, while being the biggest step (and a big > applause for everybody who contributed to this herculean task!) is > unfortunately not the completion. The bug reporting and fixing that follows > it shouldn't just be ignored.
^ This says it better than I could. -- 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 view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/CAOTD34b-%3DW%2Biah4Cd0-nhNFNu1rs9Qq095cyhRM_g_U02q62Kg%40mail.gmail.com.
