On Saturday, September 5, 2015 at 9:27:50 PM UTC+2, bluescarni wrote:
>
> With Boost Python you also get the original C++ class as a Python class.
>
Of course the downside is that you have to write your wrapper in C++. Which
is an awesomely capable language and can capture pretty much the entire
gdb has its own embedded python, so it often fails if you force sage's
libpython via LD_LIBRARY_PATH
On Friday, September 4, 2015 at 12:21:21 AM UTC+2, François wrote:
>
> After opening the firewall only test_gdb was left failing.
>
> Francois
>
--
You received this message because you are su
Port 8080 seems firewalled now, I moved it to the standard port:
http://fileserver.sagemath.org
On Friday, September 4, 2015 at 10:08:48 AM UTC+2, Jeroen Demeyer wrote:
>
> On 2015-08-08 21:22, Volker Braun wrote:
> > http://fileserver.sagemath.org:8080/
>
> Seems d
Arguably its a bug in how we use linbox; It uses random primes and we test
examples where there is a sizeable chance that it runs ouf of primes. There
are simply computations that you can't do with just machine integers, so a)
we shouldn't work with examples that are close to the limit and b) we
On Tuesday, September 1, 2015 at 10:42:20 AM UTC-4, Nathann Cohen wrote:
>
> Some time ago you sent a link toward a new webpage interface to upload
> tarballs. Is it what we should be using now? Should we update the
> developer's manual in any way?
>
Not yet, I'll post again when its finished.
OK I've updated the packages.
This is precisely why we need a self-service solution to managing tarballs,
which is what I've been working on. I for one don't have time to read 5
paragraphs of prose describing how to download and rename upstream tarballs
like in the linked ticket. And the chance
ss of file hosting or long-term storage.
On Friday, August 28, 2015 at 10:18:15 AM UTC-4, Simon King wrote:
>
> Hi Volker,
>
> On 2015-08-28, Volker Braun > wrote:
> > I'd recommend bitbucket/github/... Presumably you are using some kind of
> > version control for y
I'd recommend bitbucket/github/... Presumably you are using some kind of
version control for your files already. Github pages is an easy way to host
your documentation, too.
On Friday, August 28, 2015 at 5:15:11 AM UTC-4, Simon King wrote:
>
> Hi!
>
> It was pointed out on sage-support that
This is how exceptions are doctested in Python. Your test needs to catch
the exception if you want to test the additional output. Its obviously a
terrible ux if you rely on printing stuff before a lengthy traceback.
On Thursday, August 27, 2015 at 1:01:50 PM UTC-4, Daniel Krenn wrote:
>
> Havi
On Thursday, August 27, 2015 at 10:22:23 AM UTC-4, John Cremona wrote:
>
> How would I know that it had finished? After selecting the file it goes
> back to the "choose file" page with no apparant change and nothing
> happening.
>
You can upload multiple files, so its by design that you can cho
nd it worked for me.
On Thursday, August 27, 2015 at 9:10:02 AM UTC-4, John Cremona wrote:
>
>
>
> On 27 August 2015 at 09:16, John Cremona > wrote:
>
>>
>>
>> On 27 August 2015 at 03:42, Dan Drake
>> > wrote:
>>
>>> On Saturday, Augu
M UTC-4, Dan Drake wrote:
>
> On Saturday, August 8, 2015 at 12:22:18 PM UTC-7, Volker Braun wrote:
>>
>> In order to streamline updating third-party tarballs I've written a small
>> web app where you can directly upload them. That way you don't need to host
&g
On Monday, August 24, 2015 at 2:51:46 AM UTC-4, Jori Mäntysalo wrote:
>
> For example if M is an immutable matrix, would it give more speed to have
> something like M.__mul__(M, immutable=True)?
What would be the benefit over
M2 = M**2
M2.set_immutable()
other than a more cumbersome syntax t
On Monday, August 24, 2015 at 5:32:02 AM UTC-4, Simon King wrote:
>
> Do I understand correctly that "number of profiling samples in this
> function and its callees" roughly corresponds to the cumulated time
> that the function is spending?
>
Yes
--
You received this message because you are
Call accounting also costs time, especially for very small functions it can
be a significant fraction. Also, the Cython profile uses CPython's
profiling so I'm pretty sure accounting is only done whenever control gets
back to Python. Its unlikely to give you perfect numbers for cdef inline
stuf
On Saturday, August 22, 2015 at 1:38:01 PM UTC-4, Simon King wrote:
>
> Perhaps one could change the error message, hinting at gperftools-devel
> rather than gperftools.
>
Feel free to post a patch if it solves your problem...
--
You received this message because you are subscribed to the Go
It might be gperftools-devel; if in doubt just use your distribution's
package manager to search:
$ dnf whatprovides */libprofiler.so
gperftools-devel-2.4-1.fc22.x86_64 : Development libraries and headers for
gperftools
On Saturday, August 22, 2015 at 8:34:03 AM UTC-4, Simon King wrote:
>
>
IMHO both foo(copy=True) and foo(inplace=False) are fugly and best avoided.
They are unintuitive and not discoverable. Basically they are patterns for
an API that you can only use by constantly looking up the documentation.
If possible, methods should act in-place since that is the most versati
You could try the cython upgrade #18927, might be fixed there
A workaround is to build Sage in a directory whose name does not contain
accents.
On Friday, August 21, 2015 at 7:49:31 AM UTC-4, Anthony OLLB wrote:
>
> Hello,
>
> I attempted to compile SageMath 6.8-i686 from source on 32-bit
>
Apart from not spewing the entire log to stdout (can anybody read that
fast?), we should re-print the prompt on keyboard input thats not "y" or "n"
On Thursday, August 20, 2015 at 4:31:48 PM UTC-4, Jeroen Demeyer wrote:
>
> On 2015-08-18 10:51, Nathann Cohen wrote:
> >> Since I maintain an opt
On Tuesday, August 18, 2015 at 10:22:33 PM UTC+2, Jeroen Demeyer wrote:
>
> > I'd call that "working around bugs in our broken build system".
> No, it's called "not wanting to wait for hours until all dependencies
> are compiled".
There is no need to wait for the dependencies to compile, its th
On Tuesday, August 18, 2015 at 9:26:44 PM UTC+2, Jeroen Demeyer wrote:
>
> An obvious use-case is that you're switching branches/Sage versions and
> you just want to try to compile a single package and you don't want to
> be bothered by dependencies.
I'd call that "working around bugs in our br
I've seen that one before. Happens randomly. The doctest should be marked #
known bug
On Tuesday, August 18, 2015 at 11:07:33 AM UTC+2, Jori Mäntysalo wrote:
>
> I took the patch from http://trac.sagemath.org/ticket/18936 , compiled
> and
> ran long tests successfully. Then I installed dot2te
Whats the use case of attempting to install a package without dependencies?
I can't think of any. Whether you are developer or user, you always want to
have your dependencies in place before compiling something. The fact that
"sage -i" does not is 100% historical limitation. Now we can build pac
On Saturday, August 15, 2015 at 7:27:17 PM UTC+2, Jeroen Demeyer wrote:
>
> * Do not make a function/method "cdef" unless there is a specific
> reason, for example performance or a need of C calling convention.
>
+1
* In general I don't think that the "100% coverage rule" should be taken
> to
There isn't really enough data to decide.
* Does testing A and B provide full coverage of the helper?
* Is the helper performance-critical?
* Do you need Cython?
* If no, why is the code in that module at all? Source files in Sage are
almost always too long to comfortably read, the default sh
I made a POC at http://trac.sagemath.org/ticket/19029
On Thursday, August 13, 2015 at 8:04:52 PM UTC+2, Volker Braun wrote:
>
> Easy, the doctest displayhook should print Python integers differently.
> Say, as int(1) instead of just 1.
>
--
You received this message because you ar
http://old.files.sagemath.org/src-old/
On Friday, August 14, 2015 at 3:37:27 PM UTC+2, kcrisman wrote:
>
> Bump
>
> On Tuesday, August 4, 2015 at 4:32:28 AM UTC-4, Samuel Lelievre wrote:
>>
>> Hi all,
>>
>> what happened to
>>
>> http://www.sagemath.org/src-old/
>>
>> and where can one
"no words showed up and I just got a >" usually means that you have a stray
and unbalanced quotation mark somewhere. Post the entire output.
See also
http://doc.sagemath.org/html/en/developer/trac.html#manually-linking-your-public-key-to-your-trac-account
for how to set up the ssh key directly
Easy, the doctest displayhook should print Python integers differently.
Say, as int(1) instead of just 1.
On Thursday, August 13, 2015 at 3:20:11 PM UTC+2, Nathann Cohen wrote:
>
> Do you know how to enforce this standard? Can it be done in the buildbot?
>
>
--
You received this message bec
Doctests should cover those special branches. Without test coverage we have
no hope of testing anything.
On Thursday, August 13, 2015 at 5:53:40 PM UTC+2, Jori Mäntysalo wrote:
>
> if trivial: return 1
> return Integer(2)
>
--
You received this message because you are subscribed to the Googl
All Sage methods that return integers should return Sage integer elements.
Especially those that count something.
On Thursday, August 13, 2015 at 1:50:35 PM UTC+2, Jori Mäntysalo wrote:
>
> This might be of general interest. For example
>
> P = Poset()
> type(P.height()), type(P.width())
>
>
On Thursday, August 13, 2015 at 12:32:27 AM UTC+2, Nils Bruin wrote:
>
> Do you really just want equality and hash on FP groups to be what we have
> on the covering free presentation? GAP doesn't do that itself (and as a
> result computations there can lead to unbounded running times and outright
On Wednesday, August 12, 2015 at 10:32:49 PM UTC+2, Nils Bruin wrote:
>
> Yes, and I agree that we'd be deviating very far from the mathematical
> norm if we'd let `==` mean anything else for group elements. It also means,
> as Volker points out, that in group presentations where we don't have a
On Wednesday, August 12, 2015 at 8:03:01 PM UTC+2, vdelecroix wrote:
>
> As far as I understand, the problem is that this company (and its
> website) has the very same name as the free software
How's that not pretty common? E.g.
https://mariadb.com
https://mariadb.org
--
You received this mes
On Wednesday, August 12, 2015 at 3:29:13 PM UTC+2, Nathann Cohen wrote:
>
> > IMHO the real bug is that == is group-comparison instead of
> > presentation-comparison.
>
> P.S.: If you want to change it and make it a
> "presentation-comparison", this class should not be considered as a
> Group a
IMHO the real bug is that == is group-comparison instead of
presentation-comparison. This obviously breaks hashes, and therefor makes
associative containers useless. Getting rid of the hashes just means that
associative containers are now useless because the word problem is not
decidable. At le
[Top-posted to stop threadjacking the SymEngine post]
Just have the sage-python-library install using pip, assuming your system
has all the dependencies, is almost trivial. The real question is always
how to handle the dependencies, starting at a Fortran compiler. Also, just
to establish a base
On Monday, August 10, 2015 at 7:10:43 PM UTC+2, mmarco wrote:
>
> What I meant is that it doesn't make any sense to show a scary warning in
> the case of "encrypted but not verified" pages, but don't show any warning
> in the case of "neither encrypted nor verified" plain http pages. The
> secon
On Monday, August 10, 2015 at 5:34:49 PM UTC+2, vdelecroix wrote:
>
> Moreover, who can be a certificate authority?
There is always google if you want to know the requirements for a CA
Your proposal would result in daily "new certificate" warnings as you
browse to new web pages and/or certifica
On Monday, August 10, 2015 at 11:42:16 AM UTC+2, vdelecroix wrote:
>
> I agree with you: from a technical point of view this is stupid.
>
It is not. There is no security without the chain of trust. Maybe in a
parallel universe where everybody is so far on the autistic spectrum that
they religio
Building with homebrew in /usr/local is not supported and likely to fail.
Move /usr/local out of the way and try again.
On Sunday, August 9, 2015 at 11:53:51 PM UTC+2, Jeroen Sijsling wrote:
>
> Dear all,
>
> Today I tried to build Sage from source, for now just for the heck of it,
> but also
co wrote:
>
> I think that, since it involves sending login information, it should use
> https instead of http
>
>
>
> El sábado, 8 de agosto de 2015, 21:22:18 (UTC+2), Volker Braun escribió:
>>
>> In order to streamline updating third-party tarballs I've writ
I've removed the search-old-spkg bug/misfeature
at http://trac.sagemath.org/ticket/19004 (needs review)
Also adds useful diagnostics if the package name cannot be found:
$ sage -i python
Attempting to download package python
Error: could not find a package matching python
Try 'sage --pack
In order to streamline updating third-party tarballs I've written a small
web app where you can directly upload them. That way you don't need to host
files yourself. Plus, the files can be retrieved by sha1 so with a little
bit more scripting I won't always forget to manually copy them to the
m
) DYLD_PRINT_TO_FILE root exploit.
In any case, please file a bug with Apple to bring this to their attention
(even if you wont' get any useful information from them)
On Saturday, August 8, 2015 at 1:36:49 AM UTC+2, Juan Luis Varona wrote:
>
>
> > El 8 ago 2015, a las 0
On Friday, August 7, 2015 at 2:27:19 PM UTC+2, Juan Luis Varona wrote:
>
> sage -sh -c 'otool -L $SAGE_ROOT/local/lib/libsingular.dylib’
>
>
Your email client messed up the final backtick, it shoud be the same as all
the other ones but you have a U+2019 there.
sage -sh -c 'otool -L $SAGE_ROOT/loc
Also, whats the output of
sage -sh -c 'otool -L $SAGE_ROOT/local/lib/libsingular.dylib'
sage -sh -c 'otool -L
$SAGE_ROOT/local/lib/python2.7/site-packages/sage/matrix/matrix_mpolynomial_dense.so'
On Friday, August 7, 2015 at 1:56:39 PM UTC+2, Juan Luis Varona wrote:
>
> >
> >
> > And what d
ImportError:
dlopen(/Applications/sage-6.8/local/lib/python2.7/site-packages/sage/matrix/matrix_mpolynomial_dense.so,
2): Library not loaded: libsingular.dylib
Referenced from:
/Applications/sage-6.8/local/lib/python2.7/site-packages/sage/matrix/matrix_mpolynomial_dense.so
Reason: image not
Can you upload the entire log somewhere?
It seems that you made it to the documentation build, so ./sage should work
(probably crashes but possibly with an informative error message)
On Wednesday, August 5, 2015 at 8:55:54 PM UTC+2, Juan Luis Varona wrote:
>
> Thanks for the answer. As suggeste
Well there isn't much tooling (since one rarely wants to translate
docstrings), but its easy to programmatically extract docstrings. Then just
write them out in a pofile, translate, and change ? to look up the
translation.
--
You received this message because you are subscribed to the Google G
Since we aready have a private .ipython-* directory I'm fine with using the
profile_default in there.
On Wednesday, August 5, 2015 at 4:23:00 AM UTC+2, Kwankyu Lee wrote:
>
> Hi,
>
> As of version 6.8, Sage seems to read "profile_default" in
> $DOT_SAGE/ipython-* directory by default. As I reme
Since it is a recurring point of confusion I'll move all the old-style
spkgs so they will not be found automatically. You can still manually
install them if you specify the precise location of the archive folder.
Really, nobody wants to install ancient Sage packages.
Harald: The current achive
lding sage?
>
> On 4 August 2015 at 17:34, Volker Braun >
> wrote:
>
>> Sage can't use the ATLAS version that F22 ships with. Your compilation
>> might have failed because its difficult to tune ATLAS; You should try again:
>>
>> make distclean
>> e
EP_BUILT_SPKGS to 'yes' to prevent this.
>
> Makefile:16: recipe for target 'all' failed
> make: *** [all] Error 1
>
> I have attached my scipy log file. Running
> sudo dnf install scipy
> does not appear to solve this problem.
>
>
> On Monday, 3 Aug
On Monday, August 3, 2015 at 11:18:14 PM UTC+2, William wrote:
>
> Switching to pip for sagemath, would also involve switching to using
> standard packaging systems for things like GCC
Except on OSX where Apple is too busy rolling around in their money to
implement a standard packaging system.
On Monday, August 3, 2015 at 11:03:51 PM UTC+2, William wrote:
>
> should give an error, rather than breaking my install.
I agree, we should stop mixing old and new-style packages. Also would
simplify various scripts. Nobody needs automatic downloads of pre-git
packages. On the other hand, to
* I tried to move some of our Python packages to install using pip but at
least pip-6.1.1 inadvertently required ssl
(http://trac.sagemath.org/ticket/18131). I tracked it down to the bundled
urllib3 and worked with upstream to fix it, but haven't tried again. Or
maybe wait until openssl has its
There is no "python" package, there is only python2 and python3:
$ ls -1d build/pkgs/pyt*
build/pkgs/python2
build/pkgs/python3
On Monday, August 3, 2015 at 9:55:39 PM UTC+2, William wrote:
>
> Hi,
>
>
> (I hope you'll allow me this rant against Sage's package management,
> since I initially
Its possible that this is just a symptom of a corruption that happened
elsewhere.
The patchbot should perhaps run some clean-up on the git repo before each
build attempt to start from a known state. E.g.
git fsck --full
git stash clear
On Monday, August 3, 2015 at 8:34:25 PM UTC+2,
Agree, should be removed.
On Monday, August 3, 2015 at 2:31:06 PM UTC+2, Jan Keitel wrote:
>
> I just noticed that sage -optional lists two different TOPCOM packages,
> namely topcom and TOPCOM. Installing the former works, while the
> installation of the latter spits out the following:
>
> jkei
Fails with rm: cannot remove
'atlas-3.10.2/src/ATLAS-build/tune/blas/gemm/res': Directory not empty
Are you building on a local file system? Building on network file systems
tends to not work.
On Monday, August 3, 2015 at 10:42:59 AM UTC+2, Brenton Horne wrote:
>
> Hi,
>
> I attempted to com
Considering that this only happens on one machine its likely in the
patchbot's DOT_SAGE/ipython* configuration.
On Sunday, August 2, 2015 at 2:48:06 PM UTC+2, Frédéric Chapoton wrote:
>
> Hello,
>
> one patchbot meets the following problem with ipython:
>
> sage -t --long --warn-long 29.7 src/
On Saturday, August 1, 2015 at 4:14:33 AM UTC+2, Rob Beezer wrote:
>
> Seems only tex can understand TeX. ;-)
>
Tex it a Turing-complete language, XML is not. Hence only TeX can
understand TeX, but for XML there are various 100% compliant parsers and
converters. IMHO your decision to use XML i
The principle of least surprise would suggest that mutability should follow
the same rules independent of the matrix size. And echelon_form should
always return an immutable matrix.
On Saturday, August 1, 2015 at 1:51:12 AM UTC+2, Andrey Novoseltsev wrote:
>
> Hello,
>
> While working on http
Looks like a bug in ecl which is triggered by subtle race on
initialization. Timings are different in a debug build. It doesn't happen
on my machine, probably just because its overall faster hardware.
* Maxima doesn't do anything different when SAGE_DEBUG=yes
* Ecl compiles without optimizations
With Python being a dynamic language, you can declare anything. You need to
test whether it behaves correctly, and your example doesn't:
sage: P = Poset({1:[2,3]}, category=FiniteLatticePosets())
sage: TestSuite(P).run()
Failure in _test_not_implemented_methods:
Traceback (most recent call last
As far as UI is concerned, that would definitely be better than a random
collection of environment variables. Its still not entirely satisfactory
since we really need a way to hierarchically express options for each
package; E.g.
./configure --with-packageA-featureB=valueC
is much better s
My guess would be a resource problem: disk full / readonly filesystem / not
enough file descriptors / ...
On Wednesday, July 22, 2015 at 8:21:03 PM UTC+2, Jeroen Demeyer wrote:
>
> On 2015-07-21 17:55, rshp...@ualberta.ca wrote:
> > make[3]: *** write jobserver: Bad file descriptor. Stop.
>
I'm pretty sure you can't build Sage with clang. You'll need a fortran
compiler, too. The Sage installation guide has more.
On Friday, July 17, 2015 at 10:57:42 PM UTC+2, rshp...@ualberta.ca wrote:
>
> Debian Jessie minimal build and adding packages as needed. Possibly an
> unsatisfied depende
On Friday, July 17, 2015 at 10:25:29 AM UTC+2, Ralf Stephan wrote:
>
> 1. Can you come up with a Sage parent (other than SR) where
> element.is_zero() is sometimes undecidable?
>
The obvious example would be group rings. The word problem is undecidable
so you may not be able to figure out whether
I think that means you don't have openssl-devel so the pip built in Sage
doesn't work.
On Thursday, July 16, 2015 at 12:03:12 PM UTC+2, michele...@imtlucca.it
wrote:
>
> Exactly, Nathann, this is what I meant. Sorry if I was not clear in the
> first place.
> Anyway, that command does not work:
sage -pip install python-igraph
On Thursday, July 16, 2015 at 11:47:41 AM UTC+2, michele...@imtlucca.it
wrote:
>
> Hello!
> I am trying to improve Sage graph module: to this purpose, it would be
> excellent to add the possibility to interface with igraph, one of the most
> famous graph librarie
Looks more like a bug in OSX's dynamic linker, but you could try to compile
Sage from source if you are feeling adventurous.
On Tuesday, July 14, 2015 at 4:29:40 AM UTC+2, Juan Luis Varona wrote:
>
> I have try to install sage (the current stable version 6.7) in Mac OSX
> 11.11 (El Capitan), pu
¿¡Qué estás diciendo!?
On Monday, July 13, 2015 at 9:14:51 PM UTC+2, Dima Pasechnik wrote:
>
>
>
> On Monday, 13 July 2015 18:49:48 UTC+1, William wrote:
>>
>> On Mon, Jul 13, 2015 at 10:15 AM, Simon King
>> wrote:
>> > Hi!
>> >
>> > On 2015-07-13, Nathann Cohen wrote:
>> >> sage: sqrt
Confirmed, this is http://trac.sagemath.org/ticket/18891
Run "sage -i jmol" to fix it for now.
On Monday, July 13, 2015 at 7:14:03 PM UTC+2, rt. wrote:
>
> this is a newly built version together with gentoo, while the other
> functions works well with only 3d plot failed.
> As I excute the foll
True, but in practice its easy to have a matrix result hang off an
@cached_method.
On Monday, July 13, 2015 at 12:15:07 PM UTC+2, Simon King wrote:
>
> Hi Volker and Jori,
>
> On 2015-07-13, Volker Braun > wrote:
> > For every matrix dimension there is a parent (MatrixSp
>
> On 2015-07-13, Volker Braun > wrote:
> > You can use tor (https://www.torproject.org) to access the entire
> internet
> > or gmane (http://dir.gmane.org/gmane.comp.mathematics.sage.devel)
> > specifically for sage-devel.
>
> I wouldn't be surprised
Do you have "pdb" directory in your current working directory, or something
like that?
On Monday, July 13, 2015 at 11:12:20 AM UTC+2, Ralf Stephan wrote:
>
> Oops, rather:
> >>> from sage.all import *
> Traceback (most recent call last):
> File "", line 1, in
> File "/home/ralf/sage/local
For every matrix dimension there is a parent (MatrixSpace). You are most
likely seeing those. Use @fork / @parallel.
On Monday, July 13, 2015 at 9:44:01 AM UTC+2, Jori Mäntysalo wrote:
>
> I was asked to make a code that iterates over all matrices of given type
> and size. And once again I
Never seen that before. What do you get when you run it manually?
$ sage --python
Python 2.7.9 (default, Jul 5 2015, 10:21:02)
[GCC 5.1.1 20150618 (Red Hat 5.1.1-4)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> from sage.all import *
>>> print('IPython' in
You can use tor (https://www.torproject.org) to access the entire internet
or gmane (http://dir.gmane.org/gmane.comp.mathematics.sage.devel)
specifically for sage-devel.
Realistically, the Chinese government is always going to block mass
communication utilities that are not under their control;
Just use
export SAGE_INSTALL_GCC=yes
make
On Saturday, July 11, 2015 at 10:41:46 PM UTC+2, Luis Mailhos wrote:
>
> I have got the error twice, looks like the compiler is to blame (gcc
> version 4.8.4). I am not trying more times, seems final.
>
> Do you think it may work by installing another
On Saturday, July 11, 2015 at 2:28:40 PM UTC+2, Jeroen Demeyer wrote:
>
> Is the bug reproducible? I.e. what happens if you just try building
> again? Does the same error occur?
>
> * If yes => your compiler is broken
Or your system gcc can't work with Sage's version of ppl preloaded.
--
I've added the patch at http://trac.sagemath.org/ticket/18872, please give
it a try!
--
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...@google
Since Perl v5.21.1 according
to
https://metacpan.org/pod/release/PCM/perl-5.21.3/pod/perl5211delta.pod#A-literal-should-now-be-escaped-in-a-pattern
On Wednesday, July 8, 2015 at 7:48:10 PM UTC+2, Antonio Rojas wrote:
>
> Perl 5.22 throws warnings for unescaped '{', and pari treats warnings as
Maybe you want to be the release manager? I can assure you that running
after basic improvements of the build infrastructure it a neverending
source of great joy and time well spent.
On Sunday, July 5, 2015 at 9:02:42 PM UTC+2, Nathann Cohen wrote:
>
> > Well fine, how about we delay the next
Well fine, how about we delay the next beta until this is done.
On Sunday, July 5, 2015 at 9:00:50 PM UTC+2, Nathann Cohen wrote:
>
> > The simple fix was languishing on trac and nobody reviewed it either.
> Then I
> > refactored the code to have proper testing, its a slam-dunk. Just review
> i
The simple fix was languishing on trac and nobody reviewed it either. Then
I refactored the code to have proper testing, its a slam-dunk. Just review
it.
On Sunday, July 5, 2015 at 8:16:32 PM UTC+2, Nathann Cohen wrote:
>
> I've reviewed all blockers that I didn't fix myself. Please review the
I've reviewed all blockers that I didn't fix myself. Please review the rest:
http://trac.sagemath.org/query?priority=blocker&status=needs_review
--
You received this message because you are subscribed to the Google Groups
"sage-devel" group.
To unsubscribe from this group and stop receiving ema
The doctesting framework checks whether there are any explicit source line
numbers present; They shouldn't be in doctests as it produces fragile tests
(add a line in the source code and the test fails).
On Saturday, July 4, 2015 at 7:15:12 PM UTC+2, Simon King wrote:
>
> Hi!
>
> Working on
u are
actually aware of whats about to happen. So at the very least it needs to
warn "You are irreversibly restricting your freedom" and ask for user input.
On Friday, July 3, 2015 at 3:14:24 PM UTC+2, Jeroen Demeyer wrote:
>
> On 2015-07-03 15:10, Volker Braun wrote:
> > I
If you "convey" linked gpl and non-gpl code then you must make the source
available under terms of the gpl. Writing a script that does it silently is
still a form of conveyance on the script author's part.
On Friday, July 3, 2015 at 2:52:24 PM UTC+2, Jeroen Demeyer wrote:
>
> On 2015-07-03 11
Realistically, we don't have a chance of pickling interface objects when we
don't have any deeper understanding of the object in the interfaced system.
Unless the interfaced system supports pickling natively, which few do.
So my suggestion would be to define x.dumps() as x.sage().dumps(), that i
I'd be happy with any progress on symbolics, and Ralf did already a great
job there!
Just to spell out the obvious, software engineers are actually sought-after
by industry (unlike mathematicians, let's face it). If you want to hire a
top-notch engineer who's in it for the money then you'll ha
If they don't want their code to be in Sage then they can keep it.
--
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 post
On Friday, June 26, 2015 at 12:45:30 PM UTC+2, Simon King wrote:
>
> > If in doubt always use the minimal number of commits; There is nothing
> to
> > be gained by extra stuff in the history that you revert later.
>
> If I recall correctly, I found severeal git introductions recommending
> to c
The best solution is probably to implement your proof of concept in an
altogether separate ticket.
If in doubt always use the minimal number of commits; There is nothing to
be gained by extra stuff in the history that you revert later.
On Thursday, June 25, 2015 at 11:07:53 PM UTC+2, Simon Kin
On Tuesday, June 23, 2015 at 8:34:38 AM UTC+2, Nils Bruin wrote:
>
> Also, if A is immutable, then an explicit mutating operation would be
> uncalled for. If A is mutable then we can just implement __iadd__ without
> concern for reference counts.
>
I agree, only explicitly mutable objects are
On Tuesday, June 23, 2015 at 3:40:24 AM UTC+2, William wrote:
>
> What about something explicit, e.g.,
>A.add_inplace(B)
> which would mutate A and be very clear and explicit.
"A" might be stored in a @cached_method without the user being aware of it.
--
You received this message because
901 - 1000 of 3757 matches
Mail list logo
