On Saturday, May 27, 2023 at 4:02:27 AM UTC-7 Volker Braun wrote: NEP 29 just defines the minimum supported Python version, having a longer support window is equally NEP 29 compatible.
That was my initial interpretation too when I became interested in NEP 29 in 2020 and opened the Trac ticket https://github.com/sagemath/sage/issues/30384. For example, when we had achieved support for Python 3.6–3.9 in Sage 9.2, I wrote in the Release Tour "Hence, Sage 9.2 conforms to (and exceeds) NumPy Enhancement Proposal 29 regarding Python version support policies." (https://wiki.sagemath.org/ReleaseTours/sage-9.2#Support_for_Python_3.6.2C_3.8.2C_and_3.9_added) However, in https://github.com/sagemath/sage/issues/30384#issuecomment-1418097192 (2020-12), Tobias claimed that adopting NEP 29 means that we should (or must?) drop support for Python versions when the NEP 29 schedule says so. And this strict implementation is what Tobias's branch https://github.com/sagemath/sagetrac-mirror/compare/develop...public/python3/nep29 proposed to add into our developer's guide. (Now it's PR https://github.com/sagemath/sage/pull/35403/files) I strongly disagree with adopting this strict policy. I have little interest in discussing what the "correct interpretation" of NEP 29 is; such discussions are usually not very productive. My point is that this strict interpretation would be a poor fit for our project, as it would lead to much worse decisions on version support than what we have been doing. -- 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/ef4caddc-b72c-420d-aa60-efa084ebc14bn%40googlegroups.com.
