hich are meanwhile effectively unmaintained, though; these should
probably get moved to "untested" or "unmaintained", until someone
volunteers to further maintain them.)
-leif
On Apr 26, 2013 6:07 AM, "leif" mailto:not.rea...@online.de>> wrote:
Volke
recent versions of Clang). Its 'configure' is also broken
with these in that it doesn't "find" GMP/MPIR.
See http://trac.sagemath.org/sage_trac/ticket/13498
Note that also Sage 5.9's doctesting framework breaks its doctests.
-leif
--
() The ASCII Ribbon Camp
) so I don't even know what it
does or who the maintainer is!
There's a three years old "needs info" ticket on ACE:
http://trac.sagemath.org/sage_trac/ticket/8551
(It does have an "updated" spkg, too... :-) )
-leif
Is anybody opposed?
William
--
() The ASCII
John Cremona wrote:
Any chance of user configurable colour schemes?
Yes, please.
There's already Trac->Preferences->Syntax Highlighting where one can
choose different color schemes; haven't tried that though, and can't
tell whether it's sufficient.
-leif
On 2
Robert Bradshaw wrote:
On Sat, Apr 20, 2013 at 8:42 AM, leif wrote:
William Stein wrote:
On Sat, Apr 20, 2013 at 5:33 AM, leif mailto:not.rea...@online.de>> wrote:
People promised a couple of times (last time early this year IIRC)
that t2 (i.e., a SPARC) will come back...
t
William Stein wrote:
On Sat, Apr 20, 2013 at 5:33 AM, leif mailto:not.rea...@online.de>> wrote:
Jeroen Demeyer wrote:
On 04/20/2013 12:14 PM, Julien Puydt wrote:
Le 20/04/2013 10:29, Jeroen Demeyer a écrit :
I propose to complete remove th
naries
on hawk.
-leif
--
() The ASCII Ribbon Campaign
/\ Help Cure HTML E-Mail
--
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...@
gfortran), so obviously '-lgfortran' is
missing there.
(But it's presumably ok that this specific test fails; probably
'-lgfortran' gets added in one of the next tries; haven't investigated
the log.)
-leif
--
() The ASCII Ribbon Campaign
/\ Help Cure HTML E-Mai
-free Sage.
-leif
--
() The ASCII Ribbon Campaign
/\ Help Cure HTML E-Mail
--
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...@googl
Volker Braun wrote:
The average user will probably abort the download of obscure databases
Sounds like you're calling for a new Sage package category...
-leif
--
() The ASCII Ribbon Campaign
/\ Help Cure HTML E-Mail
--
You received this message because you are subscribed to the G
Volker Braun wrote:
In any case, if your computer can't handle large files then just don't
install huge databases.
Sure. But the average user^TM will presumably notice /after/ having
downloaded the first 4 GB... B)
-leif
--
() The ASCII Ribbon Campaign
/\ Help Cure HTML E-Ma
l ok!"
'man sem_open' would probably have been quicker... ;-)
Did you run that as root, or as the user you're building Sage with?
(And did you rebuild the Python spkg after creating the symlink?)
-leif
--
() The ASCII Ribbon Campaign
/\ Help Cure HTML E-Mail
--
You rece
Volker Braun wrote:
On Thursday, April 18, 2013 2:42:21 PM UTC+1, leif wrote:
In case we at all want to put a limit on the file size of (even "huge")
spkgs, I'd say 4 GB (for each uncompressed and untarred file as well)
would be reasonable.
Any other reason for p
le size of (even "huge")
spkgs, I'd say 4 GB (for each uncompressed and untarred file as well)
would be reasonable.
-leif
P.S.: Perhaps ask the netadmins, too...
--
() The ASCII Ribbon Campaign
/\ Help Cure HTML E-Mail
--
You received this message because you are subscribed to
pang wrote:
sudo rm -rf /dev/shm && sudo ln -s /run/shm /dev/shm
It really looked promising, but unfortunately, it didn't work: is a
reboot necessary?
I don't think so.
What does
$ ls -ld /run/shm
give? (I.e., are the permissions correct?)
-leif
--
() The ASC
27;))
2
sage: os.path.islink(os.path.join(d, 'copy.sobj'))
True
Do you agree that this is a bug?
That's not a bug, that's how [symbolic] links work.
You'd have to delete the [sym]link first, and then [re]create the file,
in order to keep the conte
Julien Puydt wrote:
Le 05/04/2013 03:28, leif a écrit :
Or otherwise ask John Cremona... ;-)
I suggest to provide him a patch adding the #include at the top of one
of the files.
Yep, but better where int64_t is actually used, i.e., in the mentioned
#else part in xmod.h, and probably some
leif wrote:
FWIW, here on Linux (included by "interface.h") pulls in
[indirectly, of course^TM], which defines int64_t et al.
The same (or, more precisely, similar) is true on Solaris.
-leif
(xmod.h is only included by arith.h, which in turn includes interface.h,
before xm
leif wrote:
Jean-Pierre Flori wrote:
On Thursday, April 4, 2013 6:22:43 PM UTC+2, Snark wrote:
Le 04/04/2013 17:52, Jean-Pierre Flori a �crit :
> I'm currently trying to build Sage on FreeBSD and the eclib spkg
failed
> because it uses int64_t in e
including cstdint in the '#else' branch shouldn't hurt
(although it's probably really a FreeBSD bug, or at least your
installation is somehow broken -- we haven't had problems on Solaris...
;-) ).
-leif
--
() The ASCII Ribbon Campaign
/\ Help Cure HTML E-Mail
--
Y
Jeroen Demeyer wrote:
On 04/01/2013 04:43 PM, leif wrote:
To me, "supported platform" doesn't mean "there exists a machine I can
build Sage on.
I think it should mean: Sage works using the *default* options, assuming
the documented prerequisites are met. If you manually set
7;s IMHO ok to rely on Sage's GCC if we encounter a
GCC version which is more or less broadly known to be broken [on
platform xy], but we shouldn't make building Sage's GCC a general
prerequisite, nor should we limit testing to just the latter case,
otherwise problems -- whethe
Jean-Pierre Flori wrote:
On Saturday, March 30, 2013 6:25:37 PM UTC+1, Jean-Pierre Flori wrote:
On Saturday, March 30, 2013 2:46:56 PM UTC+1, leif wrote:
Jean-Pierre Flori wrote:
> On Saturday, March 30, 2013 7:15:10 AM UTC+1, leif wrote:
> Just for the
? (besides disabling parallelism, which is not really a good idea
since the method can take a long time, and is trivially
parallelizable).
In any case, the maximum number of threads used should IMHO by default
be limited to SAGE_NUM_THREADS.
-leif
--
() The ASCII Ribbon Campaign
/\ Help Cure H
Jean-Pierre Flori wrote:
On Saturday, March 30, 2013 7:15:10 AM UTC+1, leif wrote:
Just for the record, I get ICEs with the GCC 4.7.2.p0 spkg (with Sun as
and ld) on Solaris SPARC (32-bit) when compiling GAP and libGAP with
-O3; -O2 works in both cases. (I didn't get them wit
more to come; haven't yet investigated further either.]
-leif
--
() The ASCII Ribbon Campaign
/\ Help Cure HTML E-Mail
--
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 i
building Sage... (Optionally:
Specify [alternate] target directory, let Sage use some
SAGE_ADDITIONAL_SPKGS_DIR[S].)
-leif
P.S.: I still think we should have some lean Sage source tarball, and
another with all standard foo stuff (iconv, patch, GCC etc.) included,
at least until we mo
David Loeffler wrote:
On 26 March 2013 12:37, leif wrote:
Volker Braun wrote:
Is it just me? I get UnicodeWarning printed during doctests on many
files. Apparently they don't cause test failures, though they might mask
failing doctests since the comparison in question is always false:
leif wrote:
Volker Braun wrote:
Is it just me? I get UnicodeWarning printed during doctests on many
files. Apparently they don't cause test failures, though they might mask
failing doctests since the comparison in question is always false:
I thought this was a known (but also harmless)
four of
these (all identical) when test[long]ing Sage 5.9.beta0.
-leif
[vbraun@localhost hg]$ sage -t sage/tests/french_book/polynomes.py
Running doctests with ID 2013-03-26-13-06-20-1ebc47c6.
Doctesting 1 file.
sage -t sage/tests/french_book/polynomes.py
/home/vbraun/opt/sage-5.9.beta1/loca
problems on FreeBSD (see some neighbour thread).
Did you use the fixed libGAP spkg from #14323 [1]?
At which point of docbuilding does it "stop"? Does it reproducibly
"stop" at the same point? Did you try running 'sage --docbuild ...' (as
opposed to 'mak
: IPython not being able to import _ctypes meanwhile
breaks a lot of doctests because of (IMHO superfluous) warnings
regarding "%gui" not being available. (This is #14309.)
Requiring _ctypes is unrelated to the purpose of the ticket, AFAIK.
-leif
--
() The ASCII Ribbon Campaign
ty in the optional
spkgs, that they all must have names like py_qt_x11-x.y.z.pn.spkg. What
do people think?
Just add periods *before* 'atlas...' as well.
-leif
--
() The ASCII Ribbon Campaign
/\ Help Cure HTML E-Mail
--
You received this message because you are subscribed
?
Don't know, but single-threaded *ptestlong*?
Did you set SAGE_NUM_THREADS=1? I'd rather use 'make testlong' instead.
-leif
--
() The ASCII Ribbon Campaign
/\ Help Cure HTML E-Mail
--
You received this message because you are subscribed to the Google Groups
&quo
e test suite is compiled with assertions
enabled.)
-leif
But
otherwise it might be nice to do, if someone knew how to do it;
zn_poly would always be able to stay a very well-supported optional
spkg.
--
() The ASCII Ribbon Campaign
/\ Help Cure HTML E-Mail
--
You received this message
John H Palmieri wrote:
On Wednesday, March 20, 2013 4:42:52 PM UTC-7, leif wrote:
Simon King wrote:
> Hi Leif,
>
> On 2013-03-20, leif > wrote:
>> We do have show(), so we could likely have play(),
>
> Why should there be a different co
1 for the other bits).
[But now it's "too late"; #14322 has positive review already.]
-leif
--
() The ASCII Ribbon Campaign
/\ Help Cure HTML E-Mail
--
You received this message because you are subscribed to the Google Groups
"sage-devel" group.
To unsubscribe
Simon King wrote:
Hi Leif,
On 2013-03-20, leif wrote:
We do have show(), so we could likely have play(),
Why should there be a different command for showing a plot, showing
a movie and showing (i.e., "showing to the ear") a sound file?
Well, you could of course exten
SAGE_PLAYER (or $SAGE_BROWSER) with a given
(audio/video/multimedia) file.
(sage/media for file formats? -- as already mentioned)
-leif
After all, I think *lab can make this, and everyting in *lab should be
developed in Sage, no?
Yours
t.
--
() The ASCII Ribbon Campaign
/\ Help Cure HTML
Nils Bruin wrote:
On Mar 20, 7:18 am, leif wrote:
And calling foo.f() will break if someone decides to no longer cache the
function or method, or to rename the attribute.
[...]
my_foo=get_uncached_function(foo)
I've seen that last argument before and I don't quite unde
my_foo=foo
# ...
[Replace my_foo by _foo or whatever seems an appropriate name... ;-)
You could of course also just overwrite foo locally to avoid changing
the function name in subsequent calls.]
This of course only works for [cached] functions you call *directly*, so
a context manager would p
usage (both positive and negative).
Add a reasonable # tol ?
-leif
I don't know
why (changes to caching?), but anyway I don't really see the point of
this test, so I propose to remove it. I have made this a *blocker* since
it causes regular doctest failures.
Removing this test ne
#SAGE_ROOT=/path/to/sage-version
-leif
On 20 March 2013 10:43, Christoph Schumacher
mailto:christoph.schumac...@web.de>> wrote:
Dear Mr Gronewald,
in the package I included with
deb http://ppa.launchpad.net/aims/sagemath/ubuntu quantal main
the path in the exec
Stephen Montgomery-Smith wrote:
On 03/19/2013 06:07 AM, leif wrote:
Jeroen Demeyer wrote:
On 2013-03-19 04:30, John H Palmieri wrote:
Is there any good way of testing whether multithreading is broken
That would of course be ideal, but we should first find out exactly
why/how it's brok
it default to SAGE_NUM_THREADS or
whatever. (The user should be able to set it as well.)
-leif
--
() The ASCII Ribbon Campaign
/\ Help Cure HTML E-Mail
--
You received this message because you are subscribed to the Google Groups
"sage-devel" group.
To unsubscribe from this grou
is a symbolic link to devel/sage-main
(unless you switched the branch, then it points to devel/sage-).
-leif
--
() The ASCII Ribbon Campaign
/\ Help Cure HTML E-Mail
--
You received this message because you are subscribed to the Google Groups
"sage-devel" group.
To unsubscribe
than...@debian.org wrote:
Am Samstag, 16. März 2013 20:57:13 UTC+1 schrieb leif:
But if we switch to git, improve Sage's package management (as a first
step, split vanilla upstream sources off the spkgs :P ), ...
Is splitting the vanilla upstream sources off planned? That would be
ent branch; 'sage -branch'
shows the [name of the] current branch; 'sage -b ' switches [back]
to branch .)
-leif
This is on sage-5.8.beta3 on Ubuntu 12.04.
Any help would be appreciated.
--
() The ASCII Ribbon Campaign
/\ Help Cure HTML E-Mail
--
You received this mess
Julien Puydt wrote:
Le 16/03/2013 18:36, leif a écrit :
Julien Puydt wrote:
Le 16/03/2013 16:57, Tobias Hansen a écrit :
== Desirable skills: ==
familiarity with shell scripts, makefiles and C libraries
You forgot python.
Autotools, C++, Common Lisp / ECL, Cython, Perl, ... ? ;-)
Do they
u PPA, ... ?
-leif
--
() The ASCII Ribbon Campaign
/\ Help Cure HTML E-Mail
--
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.
simply concatenate them with '&&'.
At least the last part should probably give an appropriate error message
telling when something (or what) went wrong:
if [ $? -ne 0 ]; then
echo >&2 "Error: Failed to copy ... to ..."
exit 1
fi
SPKG.txt looks ok to me
ng results, please don't use it, it will be removed soon".
I remember long discussions whether we should deprecate whole Sage
installations (i.e., versions) after a while ... B)
-leif
But the method still works, as it is a warning not an exception.
Deprecation has already been t
leif wrote:
Just noticed there are still subtle issues with GCC 4.7.x on Solaris
(due to its math headers).
Singular 3-1-5.p4 doesn't yet build because of these; I'll see whether
there are more instances (i.e., spkgs) later...
Nope, Singular is the only spkg, and I now rec
John Cremona wrote:
On 14 March 2013 13:42, leif wrote:
Just noticed there are still subtle issues with GCC 4.7.x on Solaris (due to
its math headers).
Singular 3-1-5.p4 doesn't yet build because of these; I'll see whether there
are more instances (i.e., spkgs) later...
(We alr
lib which had the same problem.)
-leif
--
() The ASCII Ribbon Campaign
/\ Help Cure HTML E-Mail
--
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+
ddition your
top / xload / whatever are broken...
You may try to run some lines of the tests in an interactive Sage
session and see what happens, perhaps after first running the doctests
with --verbose (although output is currently not flushed properly).
In any case, as is, 2 GB are certainly n
. If this when gets through, I'll try sending it
as a follow up.
Open a trac ticket? There G**gle won't interfere [yet].
-leif
--
() The ASCII Ribbon Campaign
/\ Help Cure HTML E-Mail
--
You received this message because you are subscribed to the Google Groups
"sage-d
devel,support}... ;-)
-leif
--
() The ASCII Ribbon Campaign
/\ Help Cure HTML E-Mail
--
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
here
else), i.e., we shouldn't require (much) more RAM for the former than
for the latter. (And IIRC you complained about Sphinx needing too much
memory for building some of the docs.)
-leif
--
() The ASCII Ribbon Campaign
/\ Help Cure HTML E-Mail
--
You received this message beca
until
a few years ago.
Well, building the Sage library apparently succeeded, which also uses
the multiprocessing module (even with SAGE_NUM_THREADS=1 or MAKE="make
-j1" I think).
-leif
--
() The ASCII Ribbon Campaign
/\ Help Cure HTML E-Mail
--
You received this message becau
ast, these would have returned ESTALE (stale NFS file handle).
syscall(0x0,0x0,0x0,0x7fbfe430,0x3fb9,0x801443e18) = 7 (0x7)
-- UNKNOWN SYSCALL 22131 --
This should be E2BIG (argument list too long)... :-)
-leif
--
() The ASCII Ribbon Campaign
/\ Help Cure HTML E-Mail
--
You rec
leif wrote:
leif wrote:
leif wrote:
sage -t --long devel/sage/sage/plot/plot3d/tachyon.py
**
File "devel/sage/sage/plot/plot3d/tachyon.py", line 490, in
sage.plot.plot3d.tachyon.Tachyon.?
Failed example:
show(t
leif wrote:
leif wrote:
sage -t --long devel/sage/sage/plot/plot3d/tachyon.py
**
File "devel/sage/sage/plot/plot3d/tachyon.py", line 490, in
sage.plot.plot3d.tachyon.Tachyon.?
Failed example:
show(t)
Expected n
leif wrote:
sage -t --long devel/sage/sage/plot/plot3d/tachyon.py
**
File "devel/sage/sage/plot/plot3d/tachyon.py", line 490, in
sage.plot.plot3d.tachyon.Tachyon.?
Failed example:
show(t)
Expected nothing
Got:
ge: t.sphere((0,-1,1), 1, 'mirror')
sage: t.sphere((2,-1,1), 0.5, 'mirror')
sage: t.sphere((2,1,1), 0.5, 'mirror')
sage: show(t)
While William's example from #7232 segfaults... (so far: because an
array index is computed from NaNs), with both the normal and the
work) when testing with '--verbose'. (With the default
ncadoctest.py, stdout and stderr aren't in sync though, so one doesn't
necessarily see /which/ test was causing it.)
Or try 'dmesg | grep tachyon' ... ;-)
-leif
--
() The ASCII Ribbon Campaign
/\ Help Cure H
John H Palmieri wrote:
On Thursday, March 7, 2013 10:05:05 AM UTC-8, leif wrote:
leif wrote:
> Julien Puydt wrote:
>> Le 02/03/2013 08:16, Julien Puydt a �crit :
>>> I'll run "make ptestlong" later, and report in this thread..
>>
leif wrote:
Julien Puydt wrote:
Le 02/03/2013 08:16, Julien Puydt a écrit :
I'll run "make ptestlong" later, and report in this thread.
- a timeout (haven't looked at it yet)
sage -t --long -force_lib
devel/sage/sage/quadratic_forms/quadratic_form__automorphisms.py
t
That's not very surprising, as testing this file doesn't only take
pretty long [now?], but also eats up *a lot* of memory (> 1 GB resident,
at least 1.8 GB virtual IIRC, on Linux x86_64 at least).
-leif
--
() The ASCII Ribbon Campaign
/\ Help Cure HTML E-Mail
--
You rece
ll require changing
doctests, but that just needs to be done once.
And people won't be able to (fully) copy-paste doctest results from an
interactive session (at least in some/more cases)... ;-)
But I don't mind.
-leif
--
() The ASCII Ribbon Campaign
/\ Help Cure HTML E-Mail
--
You r
+1 (not just for better supporting Cygwin)
We can still keep the GCC 4.6.3 spkg in optional (or archive?).
-leif
--
() The ASCII Ribbon Campaign
/\ Help Cure HTML E-Mail
--
You received this message because you are subscribed to the Google Groups
"sage-devel" group.
To unsubs
cripts.
I didn't know about that, it is very useful indeed.
'set -e' is evil, just like Python code without any 'try: ... except
...: ...'.
-leif
--
() The ASCII Ribbon Campaign
/\ Help Cure HTML E-Mail
--
You received this message because you are subscribed t
Timo Kluck wrote:
Op zondag 3 maart 2013 19:59:25 UTC+1 schreef leif het volgende:
Timo Kluck wrote:
> This target BlockDegLexOrderTest.o does not seem to be built at
all in
> the `sage -i` case, so it doesn't fail either. Is there something
in the
> en
s the "boost-cropped" spkg as well, such that
(presumably) other system-wide installations of Boost components do not
"disturb", i.e., don't get picked up by SCons. (More or less just
guessing here.) Still, the build error seems to be a version /
compatibility probl
what went wrong in Dave's first build of R, since libgcc_s.*
should be located outside the Sage tree unless Sage's GCC got installed.
-leif
--
() The ASCII Ribbon Campaign
/\ Help Cure HTML E-Mail
--
You received this message because you are subscribed to the Google Groups
&qu
rt] by
default.)
-leif
If I by any chance get a working 32 bit install, I'll try to give a 64
bits install a shot, and this should be triggered using SAGE64 :)
I'll also give a shot on Solaris 10 / sparc64, but the gcc I have there
is old and MPIR complains...
--
() The ASCII Ribbon Cam
user@frink /opt/sage/sage
$ bash -c "time true"
real0m0.000s
user0m0.000s
sys 0m0.000s
Second line means that time is a keyword of the shell
More interesting would probably be
$ which time
or
$ `which time` true
(/bin/sh is presumably almost always bash on Cygwin.)
-l
Jeroen Demeyer wrote:
If you have a Cygwin:
please tell me the output of the following shell commands:
$ time true
$ type time
$ /bin/sh -c "time true"
$ bash -c "time true"
FWIW, 'time' is (also) a bash built-in function.
-leif
--
() The ASCII Ribbon Campa
JMH wrote:
On Wednesday, February 27, 2013 12:03:40 AM UTC+1, leif wrote:
(...)
... and if you need multiple versions of the "same" package(s)
installed
at the same time (but only one of them when running Sage), can't you
just create symbolic links foo->
s in the location that
sage Python already looks in?
... and if you need multiple versions of the "same" package(s) installed
at the same time (but only one of them when running Sage), can't you
just create symbolic links foo->foo-version-x.y.z accordingly (in the
"usual&qu
from that).
[None of the scripts/programs in $SAGE_LOCAL/bin/ is supposed to be
called directly, which especially means you shouldn't put that directory
in your PATH.]
-leif
--
() The ASCII Ribbon Campaign
/\ Help Cure HTML E-Mail
--
You received this message because you are subscribe
leif wrote:
Alternatively, you could create scripts in $SAGE_LOCAL/bin/, which set
up what you need and then run Sage, but you'd have to start/run these
with './sage -c '.
Ooops, './sage --sh ' of course.
(Since $SAGE_LOCAL/bin shouldn't be in your PATH,
ons of stand-alone
executables would be even easier. For such, see
$SAGE_ROOT/devel/sage/sage/interfaces/.)
Although I agree settings for some Sage components (although most of
them "standard") are currently rather hardcoded into Sage, e.g. for R,
or matplotlib.
You'd proba
JMH wrote:
Hi,
On Sunday, February 24, 2013 11:37:47 PM UTC+1, leif wrote:
The latter should be (and IIRC used to be) the correct behaviour.
I'm not sure what you call the "correct behaviour". I can only say that
after I modified the standard "sage-env" script
feel free to fix the PIL spkg for Tcl/Tk support on
Lucid. It can hardly get more messy... ;-) )
-leif
--
() The ASCII Ribbon Campaign
/\ Help Cure HTML E-Mail
--
You received this message because you are subscribed to the Google Groups
"sage-devel" group.
To unsubscribe from
r spkg (and presumably
modify spkg/standard/deps -- some kind of Makefile -- and spkg/install,
to let Sage build and install it by default) as well.
-leif
--
() The ASCII Ribbon Campaign
/\ Help Cure HTML E-Mail
--
You received this message because you are subscribed to the Google Groups
&quo
ter" my library in a different way?
Before I source my "setup" shell script (bash), I need all Sage related
variables to be set, so the end of the standard "sage-env" script seemed
a good idea to me (I need to add some entries into "PATH",
"LD_LIBRARY_PATH&
uot;operating system" library
"/usr/lib/libpng.a", "/usr/lib/libpng.so").
Well, packages should use pkg-config (try e.g. "./sage --sh -c
'pkg-config --libs libpng'"), or [${SAGE_LOCAL}/bin/]libpng[12]-config,
which are both present... ;-)
-leif
h apparently doesn't allow ABI=32 specifically on
core[-pc-linux-gnu] (i.e., on old Intel Core CPUs, predecessors of the
Core2 processors).
(Setting ABI=standard [only] when installing the MPIR spkg is just a
work-around.)
-leif
I created a ticket at http://trac.sagemath.org/sa
ven with SAGE_CHECK off).
That was my point; I have only seen this with SAGE_CHECK=yes until David
posted.
The "extensive" test suite (run in addition if SAGE_CHECK=yes) performs
the "same" tests, but on a wider range of values.
See http://trac.sagemath.org/sage_trac/ticket
that sure right now) and (definitely) 4.7.x, on
x86/x86_64 at least. '-O2' in contrast works, as does adding '-flto'.]
-leif
--
() The ASCII Ribbon Campaign
/\ Help Cure HTML E-Mail
--
You received this message because you are subscribed to the Google Groups
"sag
; in G.lstrip(os.path.sep
> + '.'))
> or 'nodoctest' in open(G).read()[:50]):
>
> Considering what SAGE_ROOT is on my computer, all my files match the
> (os.path.sep + '.' in G.lstrip(os.path.sep + '.')) test :-)
>
> Hoping that s
gt;< build.log" or something similar)?
>
(./configure && make) 2>&1 >build.log # or | tee -a build.log
Using 'sage -i -s /path/to/new/mpir/spkg' should IMHO work as well,
provided you've built the few dependencies of MPIR in advance (e.g. by
running &
; in that directory results in "i386-apple-darwin9.8.0" instead of the
> nehalem, and I still cant run autoreconf in that directory (same reason --
> too old).
>
You probably have to explicitly configure with ABI=32. (Sage's
spkg-install does this.)
The i386
x27; spkg/logs/mpir-*.log
(The value of MPN_PATH should be on the last, i.e., 7th line -- if your
'grep' supports the context options...)
-leif
--
You received this message because you are subscribed to the Google Groups
"sage-devel" group.
To post to this group, send email
On 9 Aug., 19:58, Andrey Novoseltsev wrote:
> Ok, I've opened
>
> http://trac.sagemath.org/sage_trac/ticket/13356
>
> but it seems to me that the issue is inside ginac and I don't know how
> to fix it.
If so, report upstream... ;-)
-leif
--
--
To post to this
buntu 10.04.3], as opposed to e.g.
boxen or sage.math [still 8.04], FWIW.)
-leif
--
--
To post to this group, send an email to sage-devel@googlegroups.com
To unsubscribe from this group, send an email to
sage-devel+unsubscr...@googlegroups.com
For more options, visit this group at http://gr
you referring to 10 years old students?
Presumably rather the 30+ ones... XD
-leif
--
--
To post to this group, send an email to sage-devel@googlegroups.com
To unsubscribe from this group, send an email to
sage-devel+unsubscr...@googlegroups.com
For more options, visit this group at http://grou
27;s / xz's default compression levels, i.e., not
even using '-9' or the like.
-leif
--
--
To post to this group, send an email to sage-devel@googlegroups.com
To unsubscribe from this group, send an email to
sage-devel+unsubscr...@googlegroups.com
For more options, visit this grou
On 9 Aug., 19:24, Jeroen Demeyer wrote:
> On 2012-08-09 19:22, leif wrote:> It would probably make sense to ship newest
> versions of [m4 and]
>
> I agree with m4 but for different reasons: MPIR needs it and PPL needs
> it to build its C (not C++) interface.
That's
701 - 800 of 1271 matches
Mail list logo
