On 30/03/2014, at 8:49, leif not.rea...@online.de wrote:
Ben Salisbury wrote:
Hi all,
It seems that my dot2tex no longer functions after upgrading to 6.2.beta5.
sage: C = crystals.Tableaux(B2,shape=[2,1])
sage: view(C)
dot2tex not available. Install after running 'sage -sh'
I
You should have used make pynac has to be updated (the source of your
problem)
and sage -b won't do that.
Francois
On 2/04/2014, at 22:00, Maarten Derickx m.derickx.stud...@gmail.com wrote:
I really like the new git based workflow. It makes it much faster to update
my dev version of sage,
Same error that Karl-Dieter on OS X 10.4 see
http://trac.sagemath.org/ticket/16047
for a very minimal patch for this.
François
On 5/04/2014, at 22:08, Sébastien Labbé sla...@gmail.com wrote:
Bonjour sage-release,
On Wednesday, April 2, 2014 2:09:20 AM UTC+2, Volker Braun wrote:
Get it from
Find the PID of the process and then do strace -p PID to see what exactly it
is
doing.
Warning you'll probably get a ton of output that means nothing to you.
François
On 16/04/2014, at 23:27, John Cremona john.crem...@gmail.com wrote:
On 16 April 2014 12:18, Francois Bissey
francois.bis
in my time zone, see you after I caught some
François
On 16/04/2014, at 23:58, John Cremona john.crem...@gmail.com wrote:
On 16 April 2014 12:56, John Cremona john.crem...@gmail.com wrote:
On 16 April 2014 12:49, Francois Bissey
francois.bis...@canterbury.ac.nz wrote:
Hum, yes the parent
sage/matrix/matrix_mpolynomial_dense.so
that's who. In this case it hasn't been rebuild after nt.'s upgrade, and
Jean-Pierre
is probably right except that this ticket is in beta8 (I think). You probably
need a bit
of make distclean make
François
On 17/04/2014, at 22:54, John Cremona
Not exactly. The problem here is that some component (probably matplotlib)
detected that some gui tools are available (here tcl) and built the
corresponding
interface. If matplotlib has a gui it will try to use it by default instead of
doing everything on file.
I think we could set some default
with your sage install.
François
On 30/07/2014, at 22:21, John Cremona john.crem...@gmail.com wrote:
On 30 July 2014 11:10, Francois Bissey francois.bis...@canterbury.ac.nz
wrote:
Not exactly. The problem here is that some component (probably matplotlib)
detected that some gui tools
Hum... some configuration has gone bad on two fronts:
g++ -O2 -g -fPIC -I.. -I/Users/kabel/sage/sage/local -pipe -I. -I..
-I/Users/kabel/sage/sage/local -I/Users/kabel/sage/sage/local/include
-I/Users/kabel/sage/sage/local/include -I/Users/kabel/sage/sage/local/include
-I/usr/local/include
there is something going in there. It is
already patched for
such a problem so it must be a corner case that hasn't been encountered yet.
François
On 2/08/2014, at 21:52, Francois Bissey francoi...@canterbury.ac.nz wrote:
Hum... some configuration has gone bad on two fronts:
g++ -O2
Cannot talk for Volker. As far as I can see there are a couple of things we
would like in before
release:
* Dima pointed an obvious problem that I didn't see when I gave a positive
review here:
http://trac.sagemath.org/ticket/16735#comment:7
That probably should be fixed before release.
*
On 12/08/2014, at 21:34, leif not.rea...@online.de wrote:
On 12.08.2014 11:29, Nathann Cohen wrote:
Hell !
- Nathann Cohen [first contribution]
- Nathann Cohen
[? New incarnation?]
I don't mind reincarnating, but I would prefer it to happen *after* I die.
Well,
A dependency probably need to be adjusted. I have a failed build from scratch
on my up to date 10.9.5 mac:
running install_egg_info
Writing
/Users/fbissey/build/sage-6.4.beta5/local/lib/python2.7/lib-dynload/Python-2.7.8-py2.7.egg-info
rm
The sage-on-gentoo development ebuild pulls the develop branch, you very
much pulled the carpet from under my feet while working on a patch :)
François
On 13/10/2014, at 21:11, Volker Braun vbraun.n...@gmail.com wrote:
Thats already fixed, I'll just release a beta6 shortly.
On Monday,
That means you need java 1.7 at the minimum. Google is your friend:
http://stackoverflow.com/questions/10382929/unsupported-major-minor-version-51-0
François
On 1/11/2014, at 22:31, Jeroen Demeyer jdeme...@cage.ugent.be wrote:
Unsupported major.minor version 51.0
--
You received this
But I would definitely would like a serial log so I am sure where we are.
François
On 25/11/2014, at 18:46, Francois Bissey francois.bis...@canterbury.ac.nz
wrote:
Hum… 10.7.5, what version of clang do you have. I will inspect the log in
more depth
but I am wondering if your
Very strange. Are you sure you updated gulp to 4.55 this is part of:
https://github.com/sagemath/sage/commit/57682f249283b47071f8942544e3aaf31795cb52
François
On 28/11/2014, at 07:18, Emmanuel Charpentier emanuel.charpent...@gmail.com
wrote:
Compiles okay (I have to make doc-clean ; make,
99% sure. The C++ compiler needs some elements of C++11. I am fairly sure that
nothing below 4.2 will build gcc 4.9, but I think the barrier is either 4.5 or
4.6.
I would to check but I think I build it with 4.6 (on AIX 5.3).
François
On 5/12/2014, at 22:58, Jeroen Demeyer jdeme...@cage.ugent.be
Note that for OS X where the headers are missing I copied headers in the source.
IML is the only package I can think of requiring cblas.h we could just always
copy
the header I don’t know how it would hurt.
François
On 6/12/2014, at 00:55, Jean-Pierre Flori jpfl...@gmail.com wrote:
On
Could please people submitting logs generate them serially “-j1”.
The following line probably explains a lot of things:
/bin/bash ../libtool --tag=CC --mode=compile gcc -std=gnu99 -DHAVE_CONFIG_H
-I. -I/usr/local/sage-6.4/local/include/libpng12
-I/usr/local/sage-6.3/local/include/freetype2
Rebuilding everything serial is not necessary. You can change the MAKE command
and then do ./sage -f pkgX it will work perfectly.
In your case my suspicion is that it will be broken until
local/lib/pkgconfig/freetype2.pc
is fixed.
François
On 6/12/2014, at 09:07, Emmanuel Charpentier
Don’t remember where I got the C++11 from but Volker is correct according to
the gcc folks themselves:
https://gcc.gnu.org/install/prerequisites.html
On 6/12/2014, at 17:33, kcrisman kcris...@gmail.com wrote:
The C++ compiler needs some elements of C++11.
In theory everything should
It should compile against the vectorize library from apple. Don’t you have
/usr/lib/libclas.dylib on that machine?
Francois
On 13/12/2014, at 02:16, kcrisman kcris...@gmail.com wrote:
35ae4fd Trac #16706: Update IML to 1.0.4
Turns out something here broke compilation on some Macs.
Can you send me the config.log that would help finding what went
wrong to some extent.
François
On 13/12/2014, at 07:01, kcrisman kcris...@gmail.com wrote:
It should compile against the vectorize library from apple. Don’t you have
/usr/lib/libclas.dylib on that machine?
I can check
There had to be something I forgot…..
Changed to the right status.
François
On 14/03/2015, at 00:02, Jeroen Demeyer jdeme...@cage.ugent.be wrote:
On 2015-03-13 11:08, Francois Bissey wrote:
Done and in return I would like http://trac.sagemath.org/ticket/17946
to be reviewed
In that case
Hum… Not sure where it is stored while building but your generated Makeconf
file that is used to build R package should have a line
CC = gcc -std=gnu99
but you only seem to have CC = gcc by the look of things.
I guess we could check the effect of building R like so
CC=“gcc -std=gnu99” ./sage -i
Well, you don’t seem to have one of these path ($SAGE_LOCAL/include)
as a -I parameter so something failed at detection time.
You have:
-I/home/lmfdb/sage2/local/include/libpng12
-I/home/lmfdb/sage/local/include/freetype2
Stuff like that happened before because the wrong path was somehow
taken
It depends. To just use pip as a local installer it probably will work
fine. If you want to connect to the pip repository to install extra
python packages you’ll need SSL (and a relatively recent one too).
The pip repo just won’t let you connect without SSL.
François
On 7/04/2015, at 20:58,
I can check that which compiler and glibc were used for info?
François
On 6/06/2015, at 03:16, Jean-Pierre Flori jpfl...@gmail.com wrote:
On ppc64 running Linux I get a bunch of very bad looking errors
{{{
*** glibc detected *** python: malloc(): memory corruption:
0x010020662b90 ***
Feels unlikely as someone would have suffered from that before you would
think. I am suspecting some miscompilation and will force sage’s gcc.
François
On 23/07/2015, at 01:55, Wilfried Lübbe wlue...@gmail.com wrote:
Am Mittwoch, 22. Juli 2015 13:31:26 UTC+2 schrieb François:
sage -t
Scientific linux 7.1 x86_64:
--
sage -t --long --warn-long 314.9
src/sage/schemes/elliptic_curves/ell_rational_field.py # 2 doctests failed
sage -t --long --warn-long 314.9 src/sage/lfunctions/sympow.py # 4 doctests
failed
The palp failure disappeared in the two subsequent builds I have
done one with host gcc and one with sage’s gcc. Not sure what went wrong the
first time.
sympow’s problem has persisted all throughout.
cpuinfo:
[frb15@hp1 sage-6.8.rc1]$ tail -27 /proc/cpuinfo
processor : 111
vendor_id
It looks, to me, like /usr/local/sage/sage-1/local/bin/pkg-config was
actually a script redirecting to /usr/bin/pkg-config.
And the later seem to have disappeared.
Don’t take it wrong but it looks like you had a hack in place and
that this was done quite out of sage's build system.
I guess
Open a ticket we forgot to make it depend from pkg-config. Probably
all the machines testing it had a system instal so we didn’t notice.
François
> On 5/09/2015, at 23:04, John Cremona wrote:
>
> I also got an error with the brial package (after the same sequence as
>
And the ticket is ready for review.
François
> On 6/09/2015, at 08:36, Francois Bissey <francois.bis...@canterbury.ac.nz>
> wrote:
>
> No surprise on brial pig-config is not on OS X by default. That’s the thing
> I didn’t do while reviewing, build from scratch on
I cannot see the short log.
François
> On 6/09/2015, at 08:26, Justin C. Walker wrote:
>
> On 10.6.8 (Dual 6-core Xeons), it blew up building brial. The complaint was
> "no suitable python interpreter found". The (short) log file is attached.
> 48 spkgs built. This was
Are you moving sage or is it just after the build? I guess adding
header pad_max_install_names on OS X is not a bad idea and
would probably help Volker’s work.
François
> On 13/12/2015, at 18:51, John H Palmieri wrote:
>
> Building from scratch on OS X 10.11 fails for
Fun! Your distro is built with a gcc that is more recent than the one your sage
binary has been built with - and the libstdc++ from your distro is masked by
the one from
sage.
François
> On 18/12/2015, at 23:10, Eric Gourgoulhon wrote:
>
> Hi,
>
> I've tested
trac #19641 would be “causing” this.
> On 23/12/2015, at 10:44, Jeroen Demeyer wrote:
>
> On 2015-12-22 22:13, Volker Braun wrote:
>> You need to compile from scratch
>
> Could you elaborate a bit on why this is the case.
>
> I tried not doing "make distclean" and I
It looks like one ticket for the packages of type `script` wasn’t included then.
François
> On 23/12/2015, at 23:10, Sébastien Labbé wrote:
>
> I don't know if patchbots run make distclean but all of them report "Build
> Failed" on 7.0.beta0 with the following error:
>
>
libtinfo is part of ncurses, as I thought but not always built. There are
report of
funny things happening in ncurses and that’s probably one more.
François
> On 24/12/2015, at 10:03, Francois Bissey <francois.bis...@canterbury.ac.nz>
> wrote:
>
> Never seen that kin
ldd -r on that readline library may reveal a lack of rpath.
François
> On 24/12/2015, at 09:48, Nathann Cohen wrote:
>
>> So it looks like an ncurses problem:
>
> Hm O_o
>
> But it is using Sage's readline package, and I was compiling after a
> 'distclean'
Is libtinfo.so.5 a proper library or a symlink? And what does
nm -d libtinfo.so.5 | grep tgetflag
reports?
François
> On 24/12/2015, at 10:22, Nathann Cohen wrote:
>
>> and what is the output of ldd home/ncohen/.Sage/local/lib/libtinfo.so.5
>
> Here it is:
>
>
Do I remember that g++ 6.1 is supposed to default to c++11? It looks like that
may be
the issue here. Anyone wants to either submit a fix for brial or work around
for the spkg?
François
> On 30/05/2016, at 19:24, Johan S. R. Nielsen wrote:
>
> Hi,
>
> I'm having
trac #20545 made me meet https://github.com/sphinx-doc/sphinx/issues/2470
head on in sage-on-gentoo.
Jeroen included the bit about importing “typing” (that I had installed on the
system
to run sphinx’s tests). Very ugly failure at the start of the building of the
doc.
I’ll open a ticket for
Is now trac #20702 if someone wants to help with that.
François
> On 29/05/2016, at 19:46, Francois Bissey <francois.bis...@canterbury.ac.nz>
> wrote:
>
> trac #20545 made me meet https://github.com/sphinx-doc/sphinx/issues/2470
> head on in sage-on-gentoo.
> Jeroe
Most of that patch is not needed on linux but you should consider including
the bit of it that was in a separate PR until today (when I noticed they had
included it in the no-undefined.patch)
https://github.com/BRiAl/BRiAl/pull/5
> On 1/06/2016, at 23:26, arojas wrote:
>
>
That would mean it is not applied consistently.
Will have to dissect the logs.
> On 1/06/2016, at 22:26, arojas wrote:
>
>
> export CXXFLAGS+=' -std=gnu++98' before ./configure fixes this. Why passing
> -std=gnu++98 to the compiler command (which brial already does) is not
May you should post the associated config.log.
For singular it is usually not enough to define CPPFLAGS.
I would have gone with CXX=“g++ -std=gnu++98” there
are places in singular where CPPFLAGS/CFLAGS are not used.
François
> On 19/06/2016, at 00:37, leif wrote:
>
>
It is related to the hang. We should have worked on that after we got git back.
https://trac.sagemath.org/ticket/20845
and if you want to have more gory details and confirmation:
https://github.com/cschwan/sage-on-gentoo/issues/428
and you can try to add this patch to ecl and rebuild it:
> On 1/05/2016, at 05:27, Justin C. Walker wrote:
>
> sage -t --long --warn-long 62.2
> src/sage/schemes/elliptic_curves/ell_number_field.py # 1 doctest failed
> **
> File
Looks like “six” needs to be a build dependency of pathlib2.
Installed
/Users/Sage/sage-7.2/local/lib/python2.7/site-packages/pathlib2-2.1.0-py2.7.egg
Processing dependencies for pathlib2==2.1.0
Searching for six
Reading https://pypi.python.org/simple/six/
Download error on
OK that’s weird. I dealt with that particular doctest in the upgrade
to maxima 5.39.0 that was merged after 7.6.beta1 got out.
This is the last commit of the maxima upgrade ticket.
We were sure that was the maxima upgrade that did it. May be not in the
end.
François
> On 31/01/2017, at 20:27,
> On 9/02/2017, at 12:30, Volker Braun wrote:
>
> As always, you can get the latest beta version from the "develop" git branch.
Not yet present on github. I rely on github to push the sage-on-gentoo
equivalent.
François
--
You received this message because you are
That’s curious. eclib as only one library libec.so(.x.x) and it is not
depending on boehm-gc.
Now if you are talking about libecl.so or another component from ecl (embedded
common lisp)
that would make more sense since it depends on boehm-gc.
François
> On 17/09/2016, at 12:27,
> On 18/08/2016, at 19:38, Francois Bissey <francois.bis...@canterbury.ac.nz>
> wrote:
>
> Stack smashing on 32bits
> [dochtml] Setting permissions of DOT_SAGE directory so only you can read and
> write it.
> [dochtml] *** stack smashing detected ***: p
Stack smashing on 32bits
[dochtml] Setting permissions of DOT_SAGE directory so only you can read and
write it.
[dochtml] *** stack smashing detected ***: python terminated
[dochtml]
[dochtml]
I/We had to deal with that
OK looking closer the perl module is in pari’s sources in `src/desc`.
It looks like pari relies on “.” being part of `@INC` but it is false in your
case.
> On 2/10/2016, at 07:47, Francois Bissey <francois.bis...@canterbury.ac.nz>
> wrote:
>
> Looks like we need a sp
Looks like we need a specific perl module installed or present:
gcc -c -I. -I../src/headers -fPIC -O3 -Wall -fno-strict-aliasing
-fomit-frame-pointer -g -o emacs.o ../src/gp/emacs.c
f=funclist-$$-linux-x86_64.tmp; (cd ../src/desc && /usr/bin/perl merge_822
../../src/funclist > $f) && mv
Do you know what is the status in Gentoo? I haven’t seen anything
about it but you could have.
François
> On 2/10/2016, at 12:05, Michael Orlitzky <mich...@orlitzky.com> wrote:
>
> On 10/01/2016 03:19 PM, Francois Bissey wrote:
>> OK looking closer the perl module is in p
.
François
> On 6/11/2016, at 23:29, Francois Bissey <francois.bis...@canterbury.ac.nz>
> wrote:
>
> I don’t like the debian fix. In fact it doesn’t look like a fix at all.
> I would need a serious explanation of their change.
> That “-Wl,-r” already caused trouble in s
PIE (position independent executable - a relative of PIC)
> wrong.
>
> +1 about having severe penalties for crippling compilers.
>
> François
>
> > On 6/11/2016, at 12:19, 'Bill Hart' via sage-release
> > <sage-r...@googlegroups.com> wrote:
> >
on there, on Sierra with the latest Xcode (8.1).
François
> On 8/11/2016, at 09:50, Francois Bissey <francois.bis...@canterbury.ac.nz>
> wrote:
>
> OK I will have to give it a try on my laptop. It will be a few hours before
> I can report something.
>
> François
>
>
That was clearly a mistake in flint/arb in the first place. “-r” is a flag
for the linker, not sure what gcc was doing with it in the first place.
Has debian reported upstream?
François
> On 6/11/2016, at 07:45, Dima Pasechnik wrote:
>
>
>
> On Saturday, November 5, 2016
Hum… I was reading the patch in reverse… It changes the correct
form of “-Wl,-r” to “-r”. Probably so that it is ignored by gcc
altogether. That probably generate warnings though.
I can see why gcc complains when -pie is enabled. “-r” is supposed
to create an intermediary shared object that is
Unlikely you are running hardened. To check run:
eselect profile list
If the profile that is selected (with a *) starts with
“hardened” then yes you are using hardened. If not,
you are just normal, like me.
On that ubuntu machine can you post the output of
gcc -dumpspecs
please? I’d like to see
> On 6/11/2016, at 12:19, 'Bill Hart' via sage-release
> <sage-release@googlegroups.com> wrote:
>
> On 5 November 2016 at 20:11, Francois Bissey
> <francois.bis...@canterbury.ac.nz> wrote:
> That was clearly a mistake in flint/arb in the first place. “-r” is a flag
>
OK I will have to give it a try on my laptop. It will be a few hours before
I can report something.
François
> On 8/11/2016, at 09:46, John H Palmieri wrote:
>
> I applied #21749, but it didn't help.
>
> John
>
>
> On Sunday, November 6, 2016 at 2:46:26 PM UTC-8,
Not similar. For some reason your version of cysignals has not
been upgraded to 1.3.2 - you still have 1.1.x.
François
> On 26/11/2016, at 19:38, Peleg Michaeli wrote:
>
> I have a similar error:
>
> =
>
> [libs ] reading sources... [100%] sage/rings/pari_ring
>
./sage -p cysignals
then
make
hopefully that should be enough.
François
> On 26/11/2016, at 20:26, Peleg Michaeli wrote:
>
> What do you suggest then? make dist-clean..?
>
> On Saturday, 26 November 2016 08:42:59 UTC+2, François wrote:
> Not similar. For some reason
I am guessing it is picked up automagically by zeromq.
Sage doesn’t provide it, it was detected from your system.
I’ll look into it.
François
> On 26/11/2016, at 19:46, Steven Trogdon wrote:
>
> OK, after looking at the logs beta3 was built on 11/17. libsodium was
>
libsodium is not a sage package so it comes from Gentoo.
libsodium.so.13 was provided by libsodium-1.0.2.
All the other libsodium available in Gentoo provide
libsodium.so.18.
Would you have by chance been doing an upgrade rather
than a build of sage from scratch - and upgraded
libsodium just
Yes, zeromq has no recorded dependencies in sage, so it wasn’t
rebuilt. ./sage -f zeroqm and then make should solve your problem.
François
> On 26/11/2016, at 19:50, Steven Trogdon wrote:
>
> The beta4 build was after doing a `git pull` && `make`.
>
> On Saturday,
Informative pages about about openssl/libressl on OS X 10.11
at the time:
https://eclecticlight.co/2016/03/23/the-tls-mess-in-os-x-el-capitan/
https://docwhat.org/el-capitan-and-the-evils-of-openssl/
Note that apple recommend using its own framework “secure transport”
instead of openssl/libressl.
d.
>
> On Saturday, November 26, 2016 at 12:54:05 AM UTC-6, François wrote:
> After inspection there is a configure switch. It is always on
> in Gentoo. We could configure zeromq in sage to never depend
> on it. Ticket time?
>
> François
>
> > On 26/11
> On 18/11/2016, at 12:59, Justin C. Walker wrote:
>
> sage -t --long --warn-long 70.6 src/sage/interfaces/singular.py
We know about that one at #21865 it can even affect linux boxes.
Singular 4.1.0 should have the fix we need.
François
--
You received this message because
I think he has gcc-5.xx but I’d like a serial log to be sure which
file we are talking about.
François
> On 21/10/2016, at 22:32, Volker Braun wrote:
>
> I guess you need a c++11-capable compiler. Which gcc version is that?
>
>
>
> On Friday, October 21, 2016 at
We’ve had another report of that before. It looks like when openblas_utest
(and probably subsequent test executables) no rpath is set. If a compatible
libgfortran is not found in the normal ldpath you get that error.
We don’t set LD_LIBRARY_PATH anymore do we?
It may pay to set one to run openblas
I pushed a branch on that ticket to fix what I suspect is
the problem.
If it fixes it, I’ll push upstream since it is a problem
in their default makefile.
François
> On 12/10/2016, at 21:04, Jeroen Demeyer wrote:
>
> I created https://trac.sagemath.org/ticket/21689
>
>
> On 10/12/2016, at 22:53, Dima Pasechnik wrote:
>
> On Saturday, December 10, 2016 at 9:14:14 AM UTC, Volker Braun wrote:
> We can't remove uncommitted files from the repository because by definition
> they aren't being tracked by the repo.
>
> The only workaround would be
Looks like https://gcc.gnu.org/bugzilla/show_bug.cgi?id=71767
François
> On 7/01/2017, at 15:23, Justin C. Walker wrote:
>
>
> On Jan 6, 2017, at 18:15 , John H Palmieri wrote:
>
>> No, it's not related to gcc 4.9.3: with a fresh Sage 7.5.rc2 tarball, pynac
>> fails to build
ot do that right now, I am preparing to leave for holidays
without internet in a few hours :)
François
> On 7/01/2017, at 17:26, Ralf Stephan <gtrw...@gmail.com> wrote:
>
> So gcc-6 is needed? I'm puzzled on what code change can trigger this.
>
>
> On Sat, Jan 7, 2017, 03:31
OK the bug from bugzilla is orthogonal to the toupper/tolower issue
at hand which seems to be a clash between C++ methods and libc symbols.
May be a macro is throwing confusion. I’ll see if I can find something before
signing off for the next week.
> On 7/01/2017, at 19:34, Francois Bis
Right! I said Justin but it was your log I inspected.
> On 7/01/2017, at 19:30, John H Palmieri wrote:
>
> In case anyone has time to look at it, I posted a link to my pynac log at
> #22136.
This email may be confidential and subject to legal privilege, it may
not
Follow up. Working on a ticket. I just did a ./sage -ba to force a rebuild.
And then run a doctest:
Removing
/home/fbissey/sandbox/git-fork/sage/local/lib/python2.7/site-packages/sage-7.5.rc0-py2.7.egg-info
Writing
They are all tagged in git anyway. And so far we don’t clean the
mirror for the individual package tarballs. So everything could
be downloaded from a given git tag.
François
> On 28/03/2017, at 08:47, Volker Braun wrote:
>
> I'm fine with deleting old beta versions,
That what I would expect. Setting to blank is not the same as
unsetting it.
> On 13/04/2017, at 14:31, Adam Webb wrote:
>
> I tried setting it to blank and it failed. Once I removed it totally then it
> started building.
This email may be confidential and subject to
Can you send the config.log file from local/var/tmp/sage/build/brial-0.8.5/src/?
François
> On 13/04/2017, at 12:20, Adam Webb wrote:
>
>
> Hi,
>
> I don't know if my set-up is somehow different but I have a problem while
> building installing brial. This occurs with
Library ordering problem. What version of linux are we talking about? And in
particular what linker.
> On 24/04/2017, at 11:29, tsc...@ucdavis.edu wrote:
>
> This branch failed for me on linux at lcalc with what looks like linking
> errors. I reverted https://trac.sagemath.org/ticket/22840 and
> On 26/04/2017, at 18:11, Eric Gourgoulhon wrote:
>
> Le mardi 25 avril 2017 22:07:55 UTC+2, François a écrit :
>
> So it got me thinking that localisation could be to blame.
>
>
> It does not seem to be the reason, because on the same computer,
> sage -t --long
> On 26/04/2017, at 07:39, Sébastien Labbé wrote:
>
>
>
>
> Stupid question because I noticed stuff on the patchbot and so on.
> What happens if you do
> export LC_ALL=C
> before running the test?
>
> $ export LC_ALL=C
> $ sage -t --long src/sage/calculus/calculus.py
>
> On 12/06/2017, at 07:10, fchapot...@gmail.com wrote:
>
> t seems that this happens on several machines:
>
> Error installing package gf2x-1.1.p2
>
> See for example
>
> https://patchbot.sagemath.org/log/0/LinuxMint/18.1/x86_64/4.4.0-59-generic/rk02-math/2017-06-11%2018:40:38?short
>
And it doesn’t on my sage-on-gentoo install. I notice that graphviz
is involved in the traceback and I don’t have graphviz installed currently.
Would all the people with the failing test happen to have graphviz?
And the one who don’t, not?
François
> On 4/05/2017, at 18:23, Steven Trogdon
93 matches
Mail list logo