Hi,
On Tue, Mar 03, 2015 at 10:44:06PM +0100, Jeroen Demeyer wrote:
> On 2015-03-03 22:06, Nicolas M. Thiery wrote:
> >Any idea of what might go wrong?
> Does "docker" involve unexpected forking or multi-threading?
I don't know. In a first approximation, this should be similar to
running
Thank you. This explains that.
FYI : I had to re-build from scratch. I don't understand why... yet !
Thanks a lot !
--
Emmanuel Charpentier
Le mardi 3 mars 2015 21:41:42 UTC+1, Volker Braun a écrit :
>
> When closing a ticket, a script replaces the branch on the trac ticket
> (which might cha
On 03/04/15 12:21, Carl Robitaille wrote:
Bonjour François,
Thank you very much for your 1 character patch :-)
Trying it right now, but I have no doubt that this will fix the problem
after reading the Makefile.
Is there a bug report already filed for this?
I don't think so but I almost cert
Bonjour François,
Thank you very much for your 1 character patch :-)
Trying it right now, but I have no doubt that this will fix the problem
after reading the Makefile.
Is there a bug report already filed for this?
Carl
--
You received this message because you are subscribed to the Google Gr
You have SAGE_CHECK enabled. I know we had that failure in
sage-on-gentoo because we produce a .lo file not a .o one. I didn't think
that was happening in vanilla sage.
Try adding
https://raw.githubusercontent.com/cschwan/sage-on-gentoo/master/sci-mathematics/flint/files/flint-2.4.4-test.patch
in
I'm guessing you are hitting resource limits, e.g. compare "ipcs -l" inside
and outside of the docker container.
Afair my image passed doctests on my machine, though that machine is still
boxed up in my closet right now...
On Tuesday, March 3, 2015 at 10:06:47 PM UTC+1, Nicolas M. Thiéry wr
On 2015-03-03 18:07, Simon King wrote:
Then a metaclass will not help you, as it is not supported by Cython
extension classes.
Yes indeed, __metaclass__ doesn't do anything in Cython.
The problem I am facing in #17890 is that currently, when a Python(!)
class defines __cmp__ and _cmp_, then _
Hello,
I tried to compile sage for the first time and got an error. Following the
instructions in the log, I'm posting here about the error I got in
/home2/downloads/sage-6.5/logs/pkgs/flint-2.4.4.log
==
mkdir -p build/interfaces/test
make[4]: Entering directory
`/home2/downloads/sage-
On 2015-03-03 22:06, Nicolas M. Thiery wrote:
Any idea of what might go wrong?
Does "docker" involve unexpected forking or multi-threading?
I have seen that error before and it usually happens because there is a
duplicate process in the doctester somehow.
Also: can you please post the *actua
Dear Sage developers,
A colleague of mine (Vincent Neri, in CC) is investigating the usage
of docker to run multiple instances of the sage patchbot within
sandboxes on a cluster here. This works smoothly, except that the
doctesting framework fails mysteriously on three files, and we haven'
When closing a ticket, a script replaces the branch on the trac ticket
(which might change in the future) with its current head sha1 (i.e. always
one fixed commit).
On Tuesday, March 3, 2015 at 9:02:41 PM UTC+1, Emmanuel Charpentier wrote:
>
> Funnier and funnier : now, on a couple of another m
Actually, I did make a Sage ticket a couple of days ago:
http://trac.sagemath.org/ticket/17875
On Tuesday, March 3, 2015 at 1:13:09 PM UTC+1, Volker Braun wrote:
>
> This is due to our new setuptools (and probably the unneeded webasserts
> #16736). In any case, I reported this upstream but no S
Funnier and funnier : now, on a couple of another machines, when I 'git
trac checkout 17234', git dutifully creates the branch, but says that the
upstream branch disapearred...
I tried to look at this via the trac.sagemath.org interface : here, the
ticket is resolved/closed : do closing a ticke
Also, Python 3 will one way or the other fix your problem: the cmp()
function and the __cmp__() magic methods are no more.
On Tuesday, March 3, 2015 at 5:38:35 PM UTC+1, Jeroen Demeyer wrote:
>
> The obvious solution is to change __cmp__ to _cmp_ for every Python
> class inheriting from Element
Yes, metaclass does exactly that. E.g. add
class SageObject(object):
__metaclass__ = SageClass
Possible issues are that there are already some metaclasses used in Sage,
e.g. some of the category stuff relies on it but doesn't convey the
metaclass systematically.
Also, there are various Cyt
Hi Jeroen,
On 2015-03-03, Jeroen Demeyer wrote:
> I want to define some kind of hook on a class E which is called whenever
> another class inherits from E.
>
> So the moment somebody does
>
> class SomeOtherClass(E):
> ...
>
> I want a hook function to be called with SomeOtherClass as argum
This is more a Python question, but maybe somebody knows a solution.
I want to define some kind of hook on a class E which is called whenever
another class inherits from E.
So the moment somebody does
class SomeOtherClass(E):
...
I want a hook function to be called with SomeOtherClass as
On Wed, Feb 25, 2015 at 7:35 AM, Travis Scrimshaw wrote:
> I'm interested.
>
> Best,
> Travis
>
Unfortunately, there was very little interesting. I'm now planning to
make a date poll and get availability and use that to find good dates.
>
> On Tuesday, February 24, 2015 at 12:29:32 PM UTC-8, Wi
This is due to our new setuptools (and probably the unneeded webasserts
#16736). In any case, I reported this upstream but no Sage trac ticket yet:
https://bitbucket.org/pypa/setuptools/issue/254/the-default-python-eggs-mode-doesnt
Workaround is
chmod 700 ~/.sage/.python-eggs
On Tuesday, Ma
Hello
I have got this kind of things in tests with 6.6.b2
doctest:1224: UserWarning: /home/patchbot/.sage//.python-eggs is writable by
group/others and vulnerable to attack when used with get_resource_filename.
Consider a more secure location (set with .set_extraction_path or the
PYTHON_EGG_CA
20 matches
Mail list logo
