Re: [sage-release] Re: Sage 8.3.beta1 released

2018-05-17 Thread John Cremona
Built 8.3.beta1 from scratch (after make distclean at least after previously reported pip version issues). Two failing files both complaining that lrslib is not found: sage -t --long src/sage/game_theory/normal_form_game.py # 2 doctests failed sage -t --long

[sage-release] Re: Sage 8.3.beta1 released

2018-05-17 Thread Eric Gourgoulhon
On Ubuntu 16.04 x86_64 Xeon E5-2623 + 16 GB RAM, from a fresh git clone + pull develop, parallel (-j16) build OK and make ptestlong failed with 4 transient doctests. 3 of them are due to "Jmol failed to create file", as usual on this computer (race error?), but the 4th one is a different type:

[sage-release] Re: Sage 8.3.beta1 released

2018-05-17 Thread Sébastien Labbé
On Tuesday, May 15, 2018 at 9:24:26 PM UTC+2, Sébastien Labbé wrote: > > On Ubuntu 16.04, the command `./sage -t -p --all --long > --optional=sage,optional,external` finishes with: > > -- > sage -t --long

Re: [sage-release] Re: Sage 8.3.beta1 released

2018-05-17 Thread John Cremona
On 16 May 2018 at 22:15, Jeroen Demeyer wrote: > On 2018-05-16 17:26, Erik Bray wrote: > >> I'm not sure, but I think this might be pip-related. I think Jeroen >> mentioned something about this to me a couple weeks ago. Is it >> possible you upgraded the pip in your Sage

Re: [sage-release] Re: Sage 8.3.beta1 released

2018-05-17 Thread Vincent Delecroix
On 15/05/2018 21:24, Sébastien Labbé wrote: On Ubuntu 16.04, the command `./sage -t -p --all --long --optional=sage,optional,external` finishes with: sage -t --long src/sage/coding/code_constructions.py ** File

Re: [sage-release] Re: Sage 8.3.beta1 released

2018-05-16 Thread Jeroen Demeyer
On 2018-05-16 17:26, Erik Bray wrote: I'm not sure, but I think this might be pip-related. I think Jeroen mentioned something about this to me a couple weeks ago. Is it possible you upgraded the pip in your Sage install? Indeed, I guess you are using pip 10 by accident. It might be a good

Re: [sage-release] Re: Sage 8.3.beta1 released

2018-05-16 Thread Maarten Derickx
pip should be able to uninstall itself so sage -pip uninstall pip sage -i pip should give you the standard sage version of pip back. I did not test this, so use at your own risk. On Wednesday, 16 May 2018 17:33:50 UTC+2, John Cremona wrote: > > > > On Wed, 16 May 2018, 16:26 Erik Bray,

Re: [sage-release] Re: Sage 8.3.beta1 released

2018-05-16 Thread Erik Bray
On Wed, May 16, 2018 at 9:56 AM, John Cremona wrote: > I started to build this yesterday after pulling from trac into a place where > beta0 had already built OK. This morning I find the lines > > [scipy-0.19.1] Skipping scipy as it is not installed. > [scipy-0.19.1]

Re: [sage-release] Re: Sage 8.3.beta1 released

2018-05-16 Thread John Cremona
I started to build this yesterday after pulling from trac into a place where beta0 had already built OK. This morning I find the lines [scipy-0.19.1] Skipping scipy as it is not installed. [scipy-0.19.1] Skipping scipy as it is not installed. [scipy-0.19.1] Skipping scipy as it is not installed.

Re: [sage-release] Re: Sage 8.3.beta1 released

2018-05-16 Thread Henri Girard
Bionic 18.04 AMD x8 64 bits, compile perfect. I started from fresh git sage because upgrading from 8.2 failed. It couldn't find many libs. By the way I found my crash problem while compiling : I had to config the bios allowing warmer processing. Le 15/05/2018 à 21:24, Sébastien Labbé a écrit 

[sage-release] Re: Sage 8.3.beta1 released

2018-05-15 Thread Samuel Lelievre
Under macOS 10.10.5, after downloading the cysignals patch https://github.com/sagemath/cysignals /commit/daef81e2fc111378ccc58b3f572eb18c70753a30.patch into build/pkgs/cysignals/patches both the build+docbuild of Python2-based Sage and the build+nodocbuild of Python3-based Sage were

[sage-release] Re: Sage 8.3.beta1 released

2018-05-15 Thread Sébastien Labbé
On Ubuntu 16.04, the command `./sage -t -p --all --long --optional=sage,optional,external` finishes with: -- sage -t --long src/sage/coding/code_constructions.py # 1 doctest failed

[sage-release] Re: Sage 8.3.beta1 released

2018-05-15 Thread Dima Pasechnik
this is already fixed in https://trac.sagemath.org/ticket/25323 please review :-) On Tuesday, May 15, 2018 at 4:44:03 PM UTC+1, John H Palmieri wrote: > > For me on OS X, I get the same failures as in earlier versions (which may > be related somehow to my system). But if I then do "sage -i

[sage-release] Re: Sage 8.3.beta1 released

2018-05-15 Thread Dima Pasechnik
On Tuesday, May 15, 2018 at 4:44:03 PM UTC+1, John H Palmieri wrote: > > For me on OS X, I get the same failures as in earlier versions (which may > be related somehow to my system). But if I then do "sage -i database_gap" > and run tests again, I get failures in permgroup.py and

[sage-release] Re: Sage 8.3.beta1 released

2018-05-15 Thread John H Palmieri
For me on OS X, I get the same failures as in earlier versions (which may be related somehow to my system). But if I then do "sage -i database_gap" and run tests again, I get failures in permgroup.py and features/gap.py. For example: sage -t --long src/sage/features/gap.py

Re: [sage-release] Re: Sage 8.3.beta1 released

2018-05-15 Thread Erik Bray
Thanks for the release. On Ubuntu 14.04 I get LOTS of test timeouts that I didn't get previously. It's possible though that the system is just uder heavy load, but even re-running the tests with --failed keeps giving me timeouts for some modules. So I'm concerned about a real performance

[sage-release] Re: Sage 8.3.beta1 released

2018-05-15 Thread Emmanuel Charpentier
FWIW : on Debian testing running on Core i7 + 16 GB RAM, same results as for 8.3.beta0, i. e. one transient and two permanent errors (the same as before) : -- sage -t --long src/sage/rings/padics/padic_lattice_element.py #