[sage-devel] Re: New Python version requirement for Sage?

2020-05-27 Thread Matthias Koeppe
On Wednesday, May 27, 2020 at 3:58:28 PM UTC-7, Matthias Koeppe wrote: > > On Wednesday, May 27, 2020 at 1:08:18 PM UTC-7, Volker Braun wrote: >> >> Which python versions do we want to be compatible with now? >> > > one needs to distinguish > > A. sage-system-python > (see

Re: [sage-devel] New Python version requirement for Sage?

2020-05-27 Thread John H Palmieri
The particular ticket in question deals with matplotlib, which has Python as a dependency, so I think we can safely build with "Sage's Python 3", which should be at least 3.6. If we can accommodate older version of Python for the rest of the Sage build — and I think we can, given the current

Re: [sage-devel] New Python version requirement for Sage?

2020-05-27 Thread Michael Orlitzky
On 5/27/20 4:08 PM, Volker Braun wrote: > Which python versions do we want to be compatible with now? Please separate  > > > A)Supported Python versions for building Sage > > Note that I'm already seeing f-strings in #29547 so unless a policy is > formulated real soon the de-facto answer is

[sage-devel] Re: New Python version requirement for Sage?

2020-05-27 Thread Samuel Lelievre
I wish Sage can be made to run with any Python >= 3.6. See this ticket: - Support minimal system Python version 3.6 https://trac.sagemath.org/ticket/29033 I love f-strings and that's a good reason to drop support for any Python < 3.6. By the way, is there any way, in Python < 3.6, to import

[sage-devel] Re: New Python version requirement for Sage?

2020-05-27 Thread Matthias Koeppe
On Wednesday, May 27, 2020 at 1:08:18 PM UTC-7, Volker Braun wrote: > > Which python versions do we want to be compatible with now? Please > separate > > A) Supported Python versions for building Sage > > Note that I'm already seeing f-strings in #29547 so unless a policy is > formulated real

Re: [sage-devel] New Python version requirement for Sage?

2020-05-27 Thread François Bissey
Sage was building with py 3.6 but no python 3 lower than that a few month ago. I don’t know if it still builds with 3.6. The officially supported python that pass the doctest suite is python 3.7. 3.6 would spout a number of broken doctests (but mostly run as far as I could tell). > On

Re: [sage-devel] New Python version requirement for Sage?

2020-05-27 Thread Dima Pasechnik
On Wed, May 27, 2020 at 9:08 PM Volker Braun wrote: > > Which python versions do we want to be compatible with now? Please separate > > > A) Supported Python versions for building Sage > > Note that I'm already seeing f-strings in #29547 so unless a policy is > formulated real soon the de-facto

[sage-devel] New Python version requirement for Sage?

2020-05-27 Thread Volker Braun
Which python versions do we want to be compatible with now? Please separate A) Supported Python versions for building Sage Note that I'm already seeing f-strings in #29547 so unless a policy is formulated real soon the de-facto answer is going to be Python 3.6+ B) Supported Python versions

[sage-devel] Bug in is_line_graph()

2020-05-27 Thread David Coudert
Bernard Lidicky reported me the following issue with method is_line_graph. I don't know how to fix that and I opened ticket https://trac.sagemath.org/ticket/29740 sage: g = Graph('O{e[{}^~z`MDZBZBkXzE^') sage: g.is_line_graph()

Re: [sage-devel] Re: conditional install of optional package

2020-05-27 Thread Thierry
Hi, thanks for the answer. On Tue, May 26, 2020 at 04:01:15PM +0100, Dima Pasechnik wrote: [...] > yes, this would install everything in (assumning there > are no contradicting options > such as mpir vs gmp, no dependencies --- which is actually the biggest > problem in this, as deps can't be

Re: [sage-devel] Re: conditional install of optional package

2020-05-27 Thread Thierry
Hi, On Tue, May 26, 2020 at 08:46:42AM -0700, Matthias Koeppe wrote: > In general, to add a configure option, you can always do the following: > > ./configure $(./config.status --config) --enable-bar Nice tip, tanks ! Thierry > > Matthias > > > -- > You received this message because