Re: [sage-devel] Re: Python 3: gap interface issue

I don't know what exactly is wrong with that docker container, but certainly GAP workspace might not be present if you never ran GAP via pexpect. It is created and rotated on the fly. Does sage: gap_console() work for you, just as a sanity check? On Mon, Feb 4, 2019 at 4:24 AM Nathan Dunfield

[sage-devel] Re: Python 3: gap interface issue

On Sunday, February 3, 2019 at 7:23:52 PM UTC-6, Simon King wrote: > > Hi Nathan, > > On 2019-02-04, Nathan Dunfield > > wrote: > > On Sage 8.6 with Python 2, the following command produces the expected > > result (namely, a sage.interfaces.gap.GapElement): > > > > sage: gap('0') > > 0 >

Re: [sage-devel] Re: compiling sage failed

or, rather, you must run "configure" AFTER "make distclean" not BEFORE! On Mon, Feb 4, 2019 at 2:15 AM Dima Pasechnik wrote: > > On Mon, Feb 4, 2019 at 2:02 AM brettpim wrote: > > > > Sorry I meant I did > > > > ./configure --prefix=$SAGE_LOCAL > > > > which is set to $SAGE_ROOT/local/ > > > >

Re: [sage-devel] Re: compiling sage failed

On Mon, Feb 4, 2019 at 2:02 AM brettpim wrote: > > Sorry I meant I did > > ./configure --prefix=$SAGE_LOCAL > > which is set to $SAGE_ROOT/local/ > > I did this command because of instructions on sagemath installation > instructions the instruction says that if you just do ./configure this is

[sage-devel] Re: compiling sage failed

Sorry I meant I did ./configure --prefix=$SAGE_LOCAL which is set to $SAGE_ROOT/local/ I did this command because of instructions on sagemath installation instructions On Sunday, 3 February 2019 20:08:20 UTC-5, brettpim wrote: > > I

Re: [sage-devel] Re: Python 3: gap interface issue

On Mon, Feb 4, 2019 at 1:23 AM Simon King wrote: > > Hi Nathan, > > On 2019-02-04, Nathan Dunfield wrote: > > On Sage 8.6 with Python 2, the following command produces the expected > > result (namely, a sage.interfaces.gap.GapElement): > > > > sage: gap('0') > > 0 > > > > whereas in Sage 8.6

[sage-devel] Re: Python 3: gap interface issue

Hi Nathan, On 2019-02-04, Nathan Dunfield wrote: > On Sage 8.6 with Python 2, the following command produces the expected > result (namely, a sage.interfaces.gap.GapElement): > > sage: gap('0') > 0 > > whereas in Sage 8.6 with Python 3 I get the following error: Works for me. In Sage 8.7.beta1

Re: [sage-devel] compiling sage failed

what was the idea of doing ./configure --prefix=SAGE_LOCAL in the 1t place? It could be a bug of a sort that the value of prefix can be anything, but SAGE_LOCAL, which is also a name of environment variable used by Sage a lot... On Mon, Feb 4, 2019 at 1:08 AM brettpim wrote: > > I am running

[sage-devel] compiling sage failed

I am running Ubuntu 18.10 64bit OS I synced my $SAGE_ROOT directory with git pull then ./configure --prefix=SAGE_LOCAL then make distclean and then make My error is below. I do not see anything in $SAGE_ROOT/logs/pkgs that looks like it is a log for this error. I can attach anything

[sage-devel] Python 3: gap interface issue

On Sage 8.6 with Python 2, the following command produces the expected result (namely, a sage.interfaces.gap.GapElement): sage: gap('0') 0 whereas in Sage 8.6 with Python 3 I get the following error: sage: gap('0') ---

[sage-devel] Re: Python 3: adding object to a category

On Sunday, February 3, 2019 at 6:39:42 PM UTC-6, Simon King wrote: > > > Thanks for the pointer, following those examples solved my problem! > > Note, however, that it only adresses the question how to use a metaclass > in py3. It does not address the question how to use Sage's category >

[sage-devel] Re: Python 3: adding object to a category

Hi Nathan, On 2019-02-04, Nathan Dunfield wrote: > On Sunday, February 3, 2019 at 2:22:33 AM UTC-6, Frédéric Chapoton wrote: >> >> git grep "@add_metaclass" src/sage >> >> will give you plenty of examples on how to add metaclasses in a py3 >> compatible way >> > > Frédéric, > > Thanks for the

[sage-devel] Re: Python 3: adding object to a category

On Sunday, February 3, 2019 at 2:22:33 AM UTC-6, Frédéric Chapoton wrote: > > git grep "@add_metaclass" src/sage > > will give you plenty of examples on how to add metaclasses in a py3 > compatible way > Frédéric, Thanks for the pointer, following those examples solved my problem! Nathan --

Re: [sage-devel] Re: Location of ecl packages

On Sunday, February 3, 2019 at 11:26:00 AM UTC-8, mmarco wrote: > > Thanks for investigating that. > > I found another way to solve the problem: compile the module to a > monolithic .fas file during package instalation, and copy it to > $SAGE_LOCAL/lib/ecl/ > > Once you have the .fas file

Re: [sage-devel] Re: Location of ecl packages

Thanks for investigating that. I found another way to solve the problem: compile the module to a monolithic .fas file during package instalation, and copy it to $SAGE_LOCAL/lib/ecl/ Once you have the .fas file there, it can be loaded by anybody. It is being worked on #27188 in case you want

Re: [sage-devel] Re: Remove sagenb documentation from the reference manual?

On Tue, Jan 29, 2019 at 8:36 AM Dima Pasechnik wrote: > > So I just ran into this now that it is not in the "usual" place > http://doc.sagemath.org/html/en/reference/notebook/sagenb/notebook/interact.html#sagenb.notebook.interact.interact > > > > Now where do we find the interact documentation?

[sage-devel] Re: Python 3: adding object to a category

Hi Nathan, Am Sonntag, 3. Februar 2019 12:59:50 UTC+1 schrieb Simon King: > > Hi Nathan, > > You approach to make SnapPyNumbers be recognised as a field is not > very efficient. Better provide the category during Parent.__init__, > which is how the category framework is supposed to be used

Re: [sage-devel] Re: Location of ecl packages

the short answer: quicklisp is not for this: 12:02 -!- Irssi: #lisp: Total of 302 nicks [1 ops, 0 halfops, 0 voices, 301 normal] 12:02 -!- Channel #lisp created Sun Aug 3 22:30:32 2003 12:02 -!- Irssi: Join to #lisp was synced in 0 secs 12:03 < dimpase_> how does one install quicklisp into a

[sage-devel] Re: Python 3: adding object to a category

Hi Nathan, You approach to make SnapPyNumbers be recognised as a field is not very efficient. Better provide the category during Parent.__init__, which is how the category framework is supposed to be used anyway. Your approach: sage: from sage.misc.classcall_metaclass import ClasscallMetaclass

[sage-devel] Re: Python 3: adding object to a category

git grep "@add_metaclass" src/sage will give you plenty of examples on how to add metaclasses in a py3 compatible way F Le dimanche 3 février 2019 03:41:02 UTC+1, Nathan Dunfield a écrit : > > Taking Frédéric's advice to heart, I have begun porting SnapPy to > Sage+Python3. (We have