Re: [sage-devel] Re: AttributeError: type object 'sage.misc.lazy_import.LazyImport' has no attribute '__div__'
Ah yes, I think this might be due to the fact that __div__ is not a special slot in python3 anymore. Apparently cython 0.29.20 doesn't cover this up anymore, so probably the code discovers it doesn't have anywhere to stick the "__div__" method that's defined on LazyImport" Since 9.2 onwards doesn't need to maintain compatibility with py2, we can get rid of this __div__ method. Possibly it should be replaced with wrappers for the other div methods that have taken its place. So, while LazyImport gets abused quite a bit in sage, the bug you ran into is probably not due to that (given that the errorsite is the actual import of LazyImport) On Sunday, June 14, 2020 at 1:10:59 PM UTC-7, Isuru Fernando wrote: > > Here's the traceback. > > Isuru > > $sage -c 1+1 > SageMath version 9.1, Release Date: 2020-05-20 > Traceback (most recent call last): > File > "/home/conda/feedstock_root/build_artifacts/sagelib_1592119926566/_test_env_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehol/bin/sage-eval", > > line 4, in > from sage.all import * > File > "/home/conda/feedstock_root/build_artifacts/sagelib_1592119926566/_test_env_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehol/lib/python3.6/site-packages/sage/all.py", > > line 117, in > import sage.misc.lazy_import > File "sage/misc/lazy_import.pyx", line 1, in init sage.misc.lazy_import > (build/cythonized/sage/misc/lazy_import.c:10174) > AttributeError: type object 'sage.misc.lazy_import.LazyImport' has no > attribute '__div__' > > > -- 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/6f009e0b-b6eb-4bad-9892-d9690500de38o%40googlegroups.com.
Re: [sage-devel] Re: AttributeError: type object 'sage.misc.lazy_import.LazyImport' has no attribute '__div__'
Here's the traceback. Isuru $sage -c 1+1 SageMath version 9.1, Release Date: 2020-05-20 Traceback (most recent call last): File "/home/conda/feedstock_root/build_artifacts/sagelib_1592119926566/_test_env_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehol/bin/sage-eval", line 4, in from sage.all import * File "/home/conda/feedstock_root/build_artifacts/sagelib_1592119926566/_test_env_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehol/lib/python3.6/site-packages/sage/all.py", line 117, in import sage.misc.lazy_import File "sage/misc/lazy_import.pyx", line 1, in init sage.misc.lazy_import (build/cythonized/sage/misc/lazy_import.c:10174) AttributeError: type object 'sage.misc.lazy_import.LazyImport' has no attribute '__div__' -- 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/CA%2B01voPSsN8UzK_Y8etqJkvWfD61vJD3T2f17cwYqTG8ZAWpQw%40mail.gmail.com.
[sage-devel] Re: AttributeError: type object 'sage.misc.lazy_import.LazyImport' has no attribute '__div__'
On Sunday, June 14, 2020 at 11:30:04 AM UTC-7, Isuru Fernando wrote: > > I'm trying to package sage 9.1 for conda and I get this error with latest > Cython 0.29.20 > > If I downgrade to Cython 0.29.12, the error goes away. Any ideas? > > I think it would be good to report where the error happens. The idea of lazy_import objects is that their location is rebound to the actual object before serious use. That obviously doesn't happen here, so it's probably a mis-imported lazyimport object. This happen when you, for instance, do from import lazily_imported_object See, for instance the discussion in https://trac.sagemath.org/ticket/19628 . The truth is: the LazyImport wrappers are probably good enough for wrapping functions, but are provably not sufficient for wrapping whole objects (on which you call __div__). So in my opiniion, the fact that you find a lazyimport object on which __div__ is called, is a bug. -- 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/2b3d8ec0-8c21-46a2-92c1-7062aa573e9eo%40googlegroups.com.
[sage-devel] AttributeError: type object 'sage.misc.lazy_import.LazyImport' has no attribute '__div__'
I'm trying to package sage 9.1 for conda and I get this error with latest Cython 0.29.20 If I downgrade to Cython 0.29.12, the error goes away. Any ideas? Isuru -- 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/CA%2B01voMQt6FMpbee7ic_%2B4BFbNHf3-Qd23Eb396cZfH_P1jreQ%40mail.gmail.com.
Re: [sage-devel] Linking to libstdc++
please see https://trac.sagemath.org/ticket/29856 for a try to remove it On Sun, Jun 14, 2020 at 9:55 AM François Bissey wrote: > > +1 - it should at least be tried. > > > On 14/06/2020, at 7:55 PM, Dima Pasechnik wrote: > > > > Probably can be just removed. > > -- > 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/FD30818B-7F1C-48FD-9FEC-F40985DEEC45%40gmail.com. -- 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/CAAWYfq09q1rqzJM%3DYz6EevJ6CBLpStg07G14mtrjLkHp3FWzqg%40mail.gmail.com.
Re: [sage-devel] Linking to libstdc++
+1 - it should at least be tried. > On 14/06/2020, at 7:55 PM, Dima Pasechnik wrote: > > Probably can be just removed. -- 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/FD30818B-7F1C-48FD-9FEC-F40985DEEC45%40gmail.com.
Re: [sage-devel] Linking to libstdc++
Probably can be just removed. On Sun, 14 Jun 2020, 05:29 François Bissey, wrote: > > > > On 14/06/2020, at 4:25 PM, Isuru Fernando wrote: > > > > Yes, it was probably NTL related. > https://github.com/sagemath/sage/commit/bc333c78b0369377657452e340cb6b3f27c5cee0 > > > > Nice digging. libcsage is no more and it’s been a while. > > > > I think distutils now knows to link with the C++ compiler even though it > compiles C++ sources with a C compiler. OSX build works fine without > `libc++` being linked in explicitly, so that must mean that we are using > clang++ on OSX at the correct place, right? > > > > I would think so. > > François > > -- > 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/55AFB8E6-5605-4EB6-AE21-D84EBB0DD233%40gmail.com > . > -- 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/CAAWYfq28UNJ0cxf7yAED-EF1mA7us7KJyZaNJeUrRR2hxA%40mail.gmail.com.