[sage-devel] Re: problems building sage documentation
On Jan 7, 2008 10:55 AM, Pablo De Napoli [EMAIL PROTECTED] wrote:
I've managed to solve the problem: The package
texlive-cyrillic (in Debian/Ubuntu) is needed to build the Sage
documentation.
Thanks!!
I've made this trac #1719
http://trac.sagemath.org/sage_trac/ticket/1719
Pablo
El Tuesday 01 January 2008 21:25:35 Pablo De Napoli escribió:
When building Sage documentation
(make in devel/doc), I've got a strange message
(I quote below the output)
output
TEXINPUTS=/media/hda2/pablo.new_home/sage/sage-2.9/devel/doc-main/commontex
: python
/media/hda2/pablo.new_home/sage/sage-2.9/devel/doc-main/tools/mkhowto
--html --about html/stdabout.dat --iconserver ../icons --favicon
../icons/pyfav.png --address See ia href=\about.html\About this
document.../a/i for information on suggesting changes. --up-link
../index.html --up-title SAGE Documentation Index --global-module-index
../modindex.html --dvips-safe --dir html/ref ref/ref.tex
+++
TEXINPUTS=/media/hda2/pablo.new_home/sage/sage-2.9/devel/doc-main/ref:/medi
a/hda2/pablo.new_home/sage/sage-2.9/devel/doc-main/commontex:/media/hda2/pab
lo.new_home/sage/sage-2.9/devel/doc-main/paper-letter:/media/hda2/pablo.new_
home/sage/sage-2.9/devel/doc-main/texinputs: +++ latex ref
*** Session transcript and error messages are
in
/media/hda2/pablo.new_home/sage/sage-2.9/devel/doc-main/html/ref/ref.how.
*** Exited with status 1.
The relevant lines from the transcript are:
+++ latex ref
This is pdfTeXk, Version 3.141592-1.40.3 (Web2C 7.5.6)
%-line parsing enabled.
entering extended mode
(/media/hda2/pablo.new_home/sage/sage-2.9/devel/doc-main/ref/ref.tex
LaTeX2e 2005/12/01
Babel v3.8h and hyphenation patterns for english, usenglishmax, dumylang,
noh
yphenation, spanish, catalan, galician, spanish, catalan, galician, loaded.
(/media/hda2/pablo.new_home/sage/sage-2.9/devel/doc-main/texinputs/manual.c
ls Document Class: manual 1998/03/03 Document class (Python manual)
(/media/hda2/pablo.new_home/sage/sage-2.9/devel/doc-main/texinputs/pypaper.
sty)
(/media/hda2/pablo.new_home/sage/sage-2.9/devel/doc-main/texinputs/fancybox
.sty Style option: `fancybox' v1.3 2000/09/19 (tvz)
) (/usr/share/texmf-texlive/tex/latex/base/report.cls
Document Class: report 2005/09/16 v1.4f Standard LaTeX document class
(/usr/share/texmf-texlive/tex/latex/base/size12.clo))
(/media/hda2/pablo.new_home/sage/sage-2.9/devel/doc-main/texinputs/fancyhdr
.sty )
Using fancier footers than usual.
(/media/hda2/pablo.new_home/sage/sage-2.9/devel/doc-main/texinputs/fncychap
.sty )
Using fancy chapter headings.
(/media/hda2/pablo.new_home/sage/sage-2.9/devel/doc-main/texinputs/python.s
ty (/usr/share/texmf-texlive/tex/latex/tools/longtable.sty)
(/usr/share/texmf-texlive/tex/latex/tools/verbatim.sty)
(/usr/share/texmf-texlive/tex/latex/graphics/color.sty
(/etc/texmf/tex/latex/config/color.cfg)
(/usr/share/texmf-texlive/tex/latex/graphics/dvips.def)
(/usr/share/texmf-texlive/tex/latex/graphics/dvipsnam.def
(/usr/share/texmf-texlive/tex/latex/base/textcomp.sty
(/usr/share/texmf-texlive/tex/latex/base/ts1enc.def))
(/usr/share/texmf-texlive/tex/latex/amsmath/amsmath.sty
For additional information on amsmath, use the `?' option.
(/usr/share/texmf-texlive/tex/latex/amsmath/amstext.sty
(/usr/share/texmf-texlive/tex/latex/amsmath/amsgen.sty))
(/usr/share/texmf-texlive/tex/latex/amsmath/amsbsy.sty)
(/usr/share/texmf-texlive/tex/latex/amsmath/amsopn.sty))
(/media/hda2/pablo.new_home/sage/sage-2.9/devel/doc-main/commontex/macros.t
ex (/usr/share/texmf-texlive/tex/latex/tools/xspace.sty))
(/media/hda2/pablo.new_home/sage/sage-2.9/devel/doc-main/commontex/boilerpl
ate. tex
(/media/hda2/pablo.new_home/sage/sage-2.9/devel/doc-main/commontex/patchlev
el.t ex))
Writing index file ref.idx
(./ref.aux) (/usr/share/texmf-texlive/tex/latex/base/ts1cmr.fd)
No file OT2cmr.fd.
! LaTeX Error: This NFSS system isn't set up properly.
See the LaTeX manual or LaTeX Companion for explanation.
Type H return for immediate help.
...
l.28 \begin{document}
?
! Emergency stop.
...
l.28 \begin{document}
No pages of output.
Transcript written on ref.log.
*** Session transcript and error messages are
in
/media/hda2/pablo.new_home/sage/sage-2.9/devel/doc-main/html/ref/ref.how.
*** Exited with status 1.
make: *** [html/ref/ref.html] Error 1
/output
Any idea of what might be happening?
It seems to be a bug in my tex instalation rather than a bug in Sage
(I'm using texlive from Ubuntu Gutsy)
Pablo
--
William Stein
Associate Professor of Mathematics
University of Washington
http://wstein.org
--~--~-~--~~~---~--~~
To post to this group, send email to sage-devel@googlegroups.com
To unsubscribe from this group, send email to [EMAIL
[sage-devel] Re: Wiris -- something like the Sage notebook sort of
On Jan 8, 8:50 pm, William Stein [EMAIL PROTECTED] wrote: On Jan 7, 2008 11:42 PM, Francois [EMAIL PROTECTED] wrote: On Jan 8, 8:22 pm, Fernando Perez [EMAIL PROTECTED] wrote: On Jan 8, 2008 12:13 AM, Francois [EMAIL PROTECTED] wrote: FWIW, it worked for me under Ubuntu Gutsy with firefox... Strange it definitely just stay blank for me. I am using the plugin from the sun jdk 1.6, do you use 1.5 or 1.6 in Ubuntu? I'm using OSX 10.5.1 on a standard mac laptop, and I just get a blank screen too. Thanks I don't feel alone anymore :) I tried to change the user agent string but that just give the same result as with opera: browser disappearance. Anyway, while it is strange I clearly fill that under problem that shouldn't have to be fixed by the end user and do something else. --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~--~~~~--~~--~--~---
[sage-devel] Re: Wiris -- something like the Sage notebook sort of
On 8 Jan., 09:37, Francois [EMAIL PROTECTED] wrote: On Jan 8, 8:50 pm, William Stein [EMAIL PROTECTED] wrote: On Jan 7, 2008 11:42 PM, Francois [EMAIL PROTECTED] wrote: On Jan 8, 8:22 pm, Fernando Perez [EMAIL PROTECTED] wrote: On Jan 8, 2008 12:13 AM, Francois [EMAIL PROTECTED] wrote: FWIW, it worked for me under Ubuntu Gutsy with firefox... Strange it definitely just stay blank for me. I am using the plugin from the sun jdk 1.6, do you use 1.5 or 1.6 in Ubuntu? I'm using OSX 10.5.1 on a standard mac laptop, and I just get a blank screen too. Thanks I don't feel alone anymore :) I tried to change the user agent string but that just give the same result as with opera: browser disappearance. Anyway, while it is strange I clearly fill that under problem that shouldn't have to be fixed by the end user and do something else. It worked on Ubuntu 7.10, Firefox 2.0.0.11, Java(TM) Plug-in 1.6.0_03- b05 --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~--~~~~--~~--~--~---
[sage-devel] Re: Wiris -- something like the Sage notebook sort of
Wiris is a closed source commercial math software company that makes a web-based interface to their own custom mathematical software. They ended the discussion by telling us that their web-based interface is (going to be) much better than ours. A closed source commercial software will NEVER be better than SAGE, at least for me. I'd prefer SAGE even if I'd have to use the terminal for the rest of my life. Regards Tiziano --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~--~~~~--~~--~--~---
[sage-devel] Anyone knows about any regression in gmp-4.2.2?
Hi,
In my efforts to get sage in Gentoo I came on something that looks
like a problem in gmp 4.2.2.
On Gentoo gmp 4.2.2 is marked stable and is part of the system (needed
by gcc). On the ground
that the only patch relevant to me was the new fast gcd code and that
it was a performance patch
and not a correctness one, I decided to use my system gmp (along with
bzip2, readline, mpfr,gd,
pari [mistake since the Galois data are not included in Gentoo],
gnutls, atlas [oh yes not another
5 hours of tunning, please] and maxima - I passed on python for now).
Having a successful build after rather minimal adjustment I decided to
test it
(sage -testall) and compare the results with a regular build. Numerous
failures at
various point. First tut.tex:
sage -t tut.tex
***
***
File tut.py, line 1126:
: x = crt(2, 1, 3, 5); x
Expected:
11
Got:
-4
**
File tut.py, line 2250:
: M.T(11).charpoly('x').factor()
Expected:
(x - 285311670612) * (x - 534612)^2
Got:
x^3 - 732255212432452092*x^2 + 732255211931384496*x -
3145012477679296873951599424
**
2 items had failures:
1 of 8 in __main__.example_48
1 of 8 in __main__.example_96
***Test Failed*** 2 failures.
For whitespace errors, see the file .doctest_tut.tex
[60.2 s]
exit code: 256
and then a few more. For the first failed test putting back gmp/mpfr
in the build
solved the problem for the second putting back pari worked [of course
system
pari was built against system gmp so it probably propagated from
there]. That
solved almost all the failed test I had (the test didn't finish got
stuck in calc.py
if memory serves me correctly).
I checked that gmp was built properly with the make check provided
in gmp,
so it doesn't look like a miscompilation.
I guess it should be investigated before this version of gmp makes it
into sage.
--~--~-~--~~~---~--~~
To post to this group, send email to sage-devel@googlegroups.com
To unsubscribe from this group, send email to [EMAIL PROTECTED]
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/
-~--~~~~--~~--~--~---
[sage-devel] Re: Wiris -- something like the Sage notebook sort of
On Jan 8, 7:00 am, Ted Kosan [EMAIL PROTECTED] wrote: I like the AJAX notebook. But my position is that an AJAX-based notebook that is so good it pushes the limits of what AJAX can do would still be no where near as capable as a well-written Java Applet-based notebook. Well, I'm in general a fan of java, done many projects - but I'm more and more convinced that a huge java applet is not the best way for any webbased project. There are two things to consider: 1. it is much heavier, concerning load time/memory and so on. it depends on installing java, that's not default on win32! and 2. on the javascript side, in the future a lot will change when there are good JIT compilers. this could make js-code possibly ~10-100x faster! h --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~--~~~~--~~--~--~---
[sage-devel] Re: Anyone knows about any regression in gmp-4.2.2?
On Jan 8, 10:00 am, Francois [EMAIL PROTECTED] wrote:
Hi,
Hi Francois,
In my efforts to get sage in Gentoo I came on something that looks
like a problem in gmp 4.2.2.
On Gentoo gmp 4.2.2 is marked stable and is part of the system (needed
by gcc). On the ground
that the only patch relevant to me was the new fast gcd code and that
it was a performance patch
and not a correctness one, I decided to use my system gmp (along with
bzip2, readline, mpfr,gd,
pari [mistake since the Galois data are not included in Gentoo],
gnutls, atlas [oh yes not another
5 hours of tunning, please] and maxima - I passed on python for now).
Having a successful build after rather minimal adjustment I decided to
test it
(sage -testall) and compare the results with a regular build. Numerous
failures at
various point. First tut.tex:
sage -t tut.tex
***
***
File tut.py, line 1126:
: x = crt(2, 1, 3, 5); x
Expected:
11
Got:
-4
**
File tut.py, line 2250:
: M.T(11).charpoly('x').factor()
Expected:
(x - 285311670612) * (x - 534612)^2
Got:
x^3 - 732255212432452092*x^2 + 732255211931384496*x -
3145012477679296873951599424
**
2 items had failures:
1 of 8 in __main__.example_48
1 of 8 in __main__.example_96
***Test Failed*** 2 failures.
For whitespace errors, see the file .doctest_tut.tex
[60.2 s]
exit code: 256
and then a few more. For the first failed test putting back gmp/mpfr
in the build
solved the problem for the second putting back pari worked [of course
system
pari was built against system gmp so it probably propagated from
there]. That
solved almost all the failed test I had (the test didn't finish got
stuck in calc.py
if memory serves me correctly).
The system's Maxima ought to be at fault here.
I checked that gmp was built properly with the make check provided
in gmp,
so it doesn't look like a miscompilation.
You may have hit http://trac.sagemath.org/sage_trac/ticket/1482 since
the system gmp didn't have the improved gcd code patched in. The
default behavior is not to pass the parameter minimal, so the issues
you see won't be fixed if you use a non-patched gmp. For many people
patching the gmp with a GPLed patch making the derived work GPL only
will be a problem since many distributions ship code that depends on
GPL being LGPL.
I guess it should be investigated before this version of gmp makes it
into sage
I am confident it will be resolved. The changes gmp 4.2.1-4.2.2 were
largely config fixes and fixes for exotic platforms, so I don't expect
any problems once we remerge our patch set.
Thanks for the report.
Cheers,
Michael
--~--~-~--~~~---~--~~
To post to this group, send email to sage-devel@googlegroups.com
To unsubscribe from this group, send email to [EMAIL PROTECTED]
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/
-~--~~~~--~~--~--~---
[sage-devel] Re: Enthought mayavi2 as a library
On Jan 8, 12:26 am, Jaap Spies [EMAIL PROTECTED] wrote: Fernando Perez wrote: On Jan 7, 2008 2:09 PM, Jaap Spies [EMAIL PROTECTED] wrote: sage: /home/jaap/downloads/sage-2.9.2/local/bin/sage-sage: line 210: 4746 Segmentation fault sage-ipython -c $SAGE_STARTUP_COMMAND; $@ [EMAIL PROTECTED] sage-2.9.2]$ Any idea? Not without a gdb backtrace, no... Can you provide one please? To be honnest I've no clue how to do that. I used to do: gdb prog core but how do I proceed here? Run ./sage -gdb - do what you did above. Then on an error you should be dropped into gdb. Punch in bt, hit return a couple times until the whole backtrace is there and then post the result. Jaap Cheers, Michael --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~--~~~~--~~--~--~---
[sage-devel] Block matrices?
Is there a way to construct block matrices in SAGE? Not just the block_sum, augment and stack functions. As an example, let A, B, C, D be matrices and i want to construct a matrix like E=[[A,B],[C,D]] Such a feature would be very nice. -vgermrk- --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~--~~~~--~~--~--~---
[sage-devel] Re: Enthought mayavi2 as a library
mabshoff wrote: On Jan 8, 12:26 am, Jaap Spies [EMAIL PROTECTED] wrote: Fernando Perez wrote: On Jan 7, 2008 2:09 PM, Jaap Spies [EMAIL PROTECTED] wrote: sage: /home/jaap/downloads/sage-2.9.2/local/bin/sage-sage: line 210: 4746 Segmentation fault sage-ipython -c $SAGE_STARTUP_COMMAND; $@ [EMAIL PROTECTED] sage-2.9.2]$ Any idea? Not without a gdb backtrace, no... Can you provide one please? To be honnest I've no clue how to do that. I used to do: gdb prog core but how do I proceed here? Run ./sage -gdb - do what you did above. Then on an error you should be dropped into gdb. Punch in bt, hit return a couple times until the whole backtrace is there and then post the result. Thanks, now I remember :) sage: import IPython sage: IPython.Shell.IPShellWX().mainloop() [New Thread -1225729136 (LWP 17468)] [Detaching after fork from child process 17469.] -- | SAGE Version 2.9.2, Release Date: 2008-01-05 | | Type notebook() for the GUI, and license() for information.| -- sage: Exiting SAGE (CPU time 0m0.55s, Wall time 3m35.50s). [Thread -1225729136 (LWP 17468) exited] Closing threads... Done. sage: Exiting SAGE (CPU time 0m0.56s, Wall time 3m38.45s). Program received signal SIGSEGV, Segmentation fault. [Switching to Thread -1209112896 (LWP 17458)] 0x0064a3ec in free () from /lib/libc.so.6 (gdb) bt #0 0x0064a3ec in free () from /lib/libc.so.6 #1 0x061db266 in destroyAllCodes () from /home/jaap/downloads/sage-2.9.2/local/lib/libm4ri.so.0 #2 0x05efb187 in __pyx_pf_4sage_6matrix_17matrix_mod2_dense_free_m4ri (__pyx_self=0x0, unused=0x0) at sage/matrix/matrix_mod2_dense.c:1634 #3 0x080c51e0 in PyEval_EvalFrameEx (f=0xaea9fbc, throwflag=0) at Python/ceval.c:3548 #4 0x080c65d5 in PyEval_EvalCodeEx (co=0xb7e6c9f8, globals=0xb7e833e4, locals=0x0, args=0xaea9d84, argcount=0, kws=0xaea9d84, kwcount=0, defs=0xa74b3b8, defcount=1, closure=0x0) at Python/ceval.c:2831 #5 0x080c4a89 in PyEval_EvalFrameEx (f=0xaea9c3c, throwflag=0) at Python/ceval.c:3660 #6 0x080c65d5 in PyEval_EvalCodeEx (co=0xb7e6cc80, globals=0xb7e833e4, locals=0x0, args=0xa7ab3b4, argcount=1, kws=0xa7ab3b8, kwcount=0, defs=0x0, defcount=0, closure=0x0) at Python/ceval.c:2831 #7 0x080c4a89 in PyEval_EvalFrameEx (f=0xa7ab264, throwflag=0) at Python/ceval.c:3660 #8 0x080c65d5 in PyEval_EvalCodeEx (co=0xb7af84a0, globals=0xb7aecdfc, locals=0x0, args=0xa7abb14, argcount=2, kws=0xa7abb1c, kwcount=0, defs=0xb7a457d8, defcount=1, closure=0x0) at Python/ceval.c:2831 #9 0x080c4a89 in PyEval_EvalFrameEx (f=0xa7ab9d4, throwflag=0) at Python/ceval.c:3660 #10 0x080c65d5 in PyEval_EvalCodeEx (co=0xb7af83c8, globals=0xb7aecdfc, locals=0x0, args=0x8f46b50, argcount=2, kws=0x8f46b58, kwcount=0, defs=0xb7a457b8, defcount=1, closure=0x0) at Python/ceval.c:2831 #11 0x080c4a89 in PyEval_EvalFrameEx (f=0x8f46a0c, throwflag=0) at Python/ceval.c:3660 #12 0x080c65d5 in PyEval_EvalCodeEx (co=0xb7ac80f8, globals=0xb7ac2bdc, locals=0x0, args=0x8e6548c, argcount=1, kws=0x8e65490, kwcount=2, defs=0xb7aced98, defcount=2, closure=0x0) at Python/ceval.c:2831 #13 0x080c4a89 in PyEval_EvalFrameEx (f=0x8e65354, throwflag=0) at Python/ceval.c:3660 #14 0x080c65d5 in PyEval_EvalCodeEx (co=0xb7e6c608, globals=0xb7ebeacc, locals=0xb7ebeacc, args=0x0, argcount=0, kws=0x0, kwcount=0, defs=0x0, defcount=0, closure=0x0) at Python/ceval.c:2831 #15 0x080c6647 in PyEval_EvalCode (co=0xb7e6c608, globals=0xb7ebeacc, locals=0xb7ebeacc) at Python/ceval.c:494 #16 0x080e52d8 in PyRun_FileExFlags (fp=0x8e8e488, filename=0xbff8cb6f /home/jaap/downloads/sage-2.9.2/local/bin/sage-gdb-pythonstartup, ---Type return to continue, or q return to quit--- start=257, globals=0xb7ebeacc, locals=0xb7ebeacc, closeit=0, flags=0xbff8bdb8) at Python/pythonrun.c:1271 #17 0x080e5567 in PyRun_SimpleFileExFlags (fp=0x8e8e488, filename=0xbff8cb6f /home/jaap/downloads/sage-2.9.2/local/bin/sage-gdb-pythonstartup, closeit=0, flags=0xbff8bdb8) at Python/pythonrun.c:877 #18 0x080571a6 in Py_Main (argc=0, argv=0xbff8be84) at Modules/main.c:134 #19 0x08056432 in main (argc=99635103, argv=0x5efb170) at ./Modules/python.c:23 (gdb) Hope this helps, Jaap --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~--~~~~--~~--~--~---
[sage-devel] Re: Wiris -- something like the Sage notebook sort of
I think that the Wiris web interface is specialized so much that it may never be better than the Sage Notebook. I currently primary use the Sage Notebook to teach a ten year old how to program with Python. Therefore, for me organization and revision control are far more important than fancy formula tools. The most important improvement to the Sage Notebook for my teaching would be far more updates in revision control as in Google Docs. I have evaluated various cells then easily lost the last changes that were evaluated. If I were designing a Sage Notebook 3 I would want detailed surveys of Google Docs and Mathematica's notebook system. I would then write an plan for cloning those two systems and combining them wile also focusing on the goals of being very fast, AJAX based, and highly customizable like Drupal. On Jan 8, 2:16 am, Harald Schilly [EMAIL PROTECTED] wrote: On Jan 8, 7:00 am, Ted Kosan [EMAIL PROTECTED] wrote: I like the AJAX notebook. But my position is that an AJAX-based notebook that is so good it pushes the limits of what AJAX can do would still be no where near as capable as a well-written Java Applet-based notebook. Well, I'm in general a fan of java, done many projects - but I'm more and more convinced that a huge java applet is not the best way for any webbased project. There are two things to consider: 1. it is much heavier, concerning load time/memory and so on. it depends on installing java, that's not default on win32! and 2. on the javascript side, in the future a lot will change when there are good JIT compilers. this could make js-code possibly ~10-100x faster! h --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~--~~~~--~~--~--~---
[sage-devel] Re: SAGE 2.10.alpha0 released
Works for me. That means #1553 is solved, though the larger SCons issue remains in case we bring in any more software that uses it. On my box (again, that's 64-bit RHEL5), sage -testall passes all tests. Kiran On Jan 7, 9:30 pm, mabshoff [EMAIL PROTECTED] dortmund.de wrote: Hello folks, and a new merge cycle has been opened. Every time I finish one they pull me back in again ;) Anyways, tarball is in http://sage.math.washington.edu/home/mabshoff/release-cycles-2.10/ Inside that directory you see another directory alpha0 which contains all updated spkgs as well as all patches. So that way you can keep an eye on development and just grab patches and spkgs as they get merged. Once I start alpha1 expect that directory to appear there, too. As usual I will ask people to do reviews and fix open bugs. Come on over to the tracker and do your best. alpha0 was mostly about mem leak fixes, I did discover a whole bunch of new ones when I did an audit with the help of valgrind of 2.9.3 and with the help of various people (see below) we fixed a whole bunch of them. We are getting to the point where the number of known leaks is going toward zero, the main culprit at the moment is Givaro. As usual let us know about any issue you encounter. I am catching some sleep now - see you in the morning. Cheers, Michael Merged in alpha0: #1092: Willem Jan Palenstijn, Michael Abshoff: small memleaks exposed by ntl_ZZ_pE #1093: Michael Abshoff, Willem Jan Palenstijn: small memleaks exposed by ntl_ZZ_pX.py #1541: Burcin Erocal: improve PolyBoRi integration #1544: David Joyner, Rich Morin: SAGE Tutorial nits #2 #1553: Kiran Kedlaya, Michael Abshoff: SCons related build failure of PolyBoRi on 64-bit RHEL5 #1598: Martin Albrecht: fix SIGSEGV in libSINGULAR interface on Solaris #1694: Michael Abshoff, Bill Hart: Update FLINT to 1.05 release #1701: Martin Albrecht: attempt to clean up currRing if deallocated #1702: Martin Albrecht: fix memleak in fplll.pyx #1703: Michael Abshoff: memleak in Singular: one mpz is leaked in longrat.cc --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~--~~~~--~~--~--~---
[sage-devel] Re: Anyone knows about any regression in gmp-4.2.2?
Certainly the first example is not exactly a bug since the answer is a
priori only determined modulo 15, so 11 is as good as -4. It's only a
bug if the solution is supposed to be guaranteed minimal, which is
what the earlier business with extended gcd was all about.
John
On 08/01/2008, mabshoff
[EMAIL PROTECTED] wrote:
On Jan 8, 10:00 am, Francois [EMAIL PROTECTED] wrote:
Hi,
Hi Francois,
In my efforts to get sage in Gentoo I came on something that looks
like a problem in gmp 4.2.2.
On Gentoo gmp 4.2.2 is marked stable and is part of the system (needed
by gcc). On the ground
that the only patch relevant to me was the new fast gcd code and that
it was a performance patch
and not a correctness one, I decided to use my system gmp (along with
bzip2, readline, mpfr,gd,
pari [mistake since the Galois data are not included in Gentoo],
gnutls, atlas [oh yes not another
5 hours of tunning, please] and maxima - I passed on python for now).
Having a successful build after rather minimal adjustment I decided to
test it
(sage -testall) and compare the results with a regular build. Numerous
failures at
various point. First tut.tex:
sage -t tut.tex
***
***
File tut.py, line 1126:
: x = crt(2, 1, 3, 5); x
Expected:
11
Got:
-4
**
File tut.py, line 2250:
: M.T(11).charpoly('x').factor()
Expected:
(x - 285311670612) * (x - 534612)^2
Got:
x^3 - 732255212432452092*x^2 + 732255211931384496*x -
3145012477679296873951599424
**
2 items had failures:
1 of 8 in __main__.example_48
1 of 8 in __main__.example_96
***Test Failed*** 2 failures.
For whitespace errors, see the file .doctest_tut.tex
[60.2 s]
exit code: 256
and then a few more. For the first failed test putting back gmp/mpfr
in the build
solved the problem for the second putting back pari worked [of course
system
pari was built against system gmp so it probably propagated from
there]. That
solved almost all the failed test I had (the test didn't finish got
stuck in calc.py
if memory serves me correctly).
The system's Maxima ought to be at fault here.
I checked that gmp was built properly with the make check provided
in gmp,
so it doesn't look like a miscompilation.
You may have hit http://trac.sagemath.org/sage_trac/ticket/1482 since
the system gmp didn't have the improved gcd code patched in. The
default behavior is not to pass the parameter minimal, so the issues
you see won't be fixed if you use a non-patched gmp. For many people
patching the gmp with a GPLed patch making the derived work GPL only
will be a problem since many distributions ship code that depends on
GPL being LGPL.
I guess it should be investigated before this version of gmp makes it
into sage
I am confident it will be resolved. The changes gmp 4.2.1-4.2.2 were
largely config fixes and fixes for exotic platforms, so I don't expect
any problems once we remerge our patch set.
Thanks for the report.
Cheers,
Michael
--
John Cremona
--~--~-~--~~~---~--~~
To post to this group, send email to sage-devel@googlegroups.com
To unsubscribe from this group, send email to [EMAIL PROTECTED]
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/
-~--~~~~--~~--~--~---
[sage-devel] Re: Enthought mayavi2 as a library
On Jan 8, 1:40 pm, Jaap Spies [EMAIL PROTECTED] wrote: mabshoff wrote: On Jan 8, 12:26 am, Jaap Spies [EMAIL PROTECTED] wrote: Fernando Perez wrote: On Jan 7, 2008 2:09 PM, Jaap Spies [EMAIL PROTECTED] wrote: sage: /home/jaap/downloads/sage-2.9.2/local/bin/sage-sage: line 210: 4746 Segmentation fault sage-ipython -c $SAGE_STARTUP_COMMAND; $@ [EMAIL PROTECTED] sage-2.9.2]$ Any idea? Not without a gdb backtrace, no... Can you provide one please? To be honnest I've no clue how to do that. I used to do: gdb prog core but how do I proceed here? Run ./sage -gdb - do what you did above. Then on an error you should be dropped into gdb. Punch in bt, hit return a couple times until the whole backtrace is there and then post the result. Thanks, now I remember :) I always thought you had forgotten more about computers than most of us know ;) sage: import IPython sage: IPython.Shell.IPShellWX().mainloop() [New Thread -1225729136 (LWP 17468)] [Detaching after fork from child process 17469.] -- | SAGE Version 2.9.2, Release Date: 2008-01-05 | | Type notebook() for the GUI, and license() for information.| -- sage: Exiting SAGE (CPU time 0m0.55s, Wall time 3m35.50s). [Thread -1225729136 (LWP 17468) exited] Closing threads... Done. sage: Exiting SAGE (CPU time 0m0.56s, Wall time 3m38.45s). Program received signal SIGSEGV, Segmentation fault. [Switching to Thread -1209112896 (LWP 17458)] 0x0064a3ec in free () from /lib/libc.so.6 (gdb) bt #0 0x0064a3ec in free () from /lib/libc.so.6 #1 0x061db266 in destroyAllCodes () from /home/jaap/downloads/sage-2.9.2/local/lib/libm4ri.so.0 #2 0x05efb187 in __pyx_pf_4sage_6matrix_17matrix_mod2_dense_free_m4ri (__pyx_self=0x0, unused=0x0) at sage/matrix/matrix_mod2_dense.c:1634 Ok, the culprit points to a cleanup function of m4ri, which under normal conditions is called only once. I am not seeing the above issue under valgrdind with pure Sage 2.10.alpha0, but as I just learned yesterday problematic free() calls might be hidden by other bugs. Imagine my surprise that once I fixed #1092 or #1093 some invalid free popped up. We currently have a problem in tp_new (see #1337) which is the only culprit on our end that could hide that. But the question remains if IPython does something funny with threads and somehow triggers more than one cleanup call. We could add some check for NULL inside destroyAllCodes(). So, anybody with a better knowledge of IPython: It seems that threads are involved, so could this somehow be triggered? #3 0x080c51e0 in PyEval_EvalFrameEx (f=0xaea9fbc, throwflag=0) at Python/ceval.c:3548 #4 0x080c65d5 in PyEval_EvalCodeEx (co=0xb7e6c9f8, globals=0xb7e833e4, locals=0x0, args=0xaea9d84, argcount=0, kws=0xaea9d84, kwcount=0, defs=0xa74b3b8, defcount=1, closure=0x0) at Python/ceval.c:2831 #5 0x080c4a89 in PyEval_EvalFrameEx (f=0xaea9c3c, throwflag=0) at Python/ceval.c:3660 #6 0x080c65d5 in PyEval_EvalCodeEx (co=0xb7e6cc80, globals=0xb7e833e4, locals=0x0, args=0xa7ab3b4, argcount=1, kws=0xa7ab3b8, kwcount=0, defs=0x0, defcount=0, closure=0x0) at Python/ceval.c:2831 #7 0x080c4a89 in PyEval_EvalFrameEx (f=0xa7ab264, throwflag=0) at Python/ceval.c:3660 #8 0x080c65d5 in PyEval_EvalCodeEx (co=0xb7af84a0, globals=0xb7aecdfc, locals=0x0, args=0xa7abb14, argcount=2, kws=0xa7abb1c, kwcount=0, defs=0xb7a457d8, defcount=1, closure=0x0) at Python/ceval.c:2831 #9 0x080c4a89 in PyEval_EvalFrameEx (f=0xa7ab9d4, throwflag=0) at Python/ceval.c:3660 #10 0x080c65d5 in PyEval_EvalCodeEx (co=0xb7af83c8, globals=0xb7aecdfc, locals=0x0, args=0x8f46b50, argcount=2, kws=0x8f46b58, kwcount=0, defs=0xb7a457b8, defcount=1, closure=0x0) at Python/ceval.c:2831 #11 0x080c4a89 in PyEval_EvalFrameEx (f=0x8f46a0c, throwflag=0) at Python/ceval.c:3660 #12 0x080c65d5 in PyEval_EvalCodeEx (co=0xb7ac80f8, globals=0xb7ac2bdc, locals=0x0, args=0x8e6548c, argcount=1, kws=0x8e65490, kwcount=2, defs=0xb7aced98, defcount=2, closure=0x0) at Python/ceval.c:2831 #13 0x080c4a89 in PyEval_EvalFrameEx (f=0x8e65354, throwflag=0) at Python/ceval.c:3660 #14 0x080c65d5 in PyEval_EvalCodeEx (co=0xb7e6c608, globals=0xb7ebeacc, locals=0xb7ebeacc, args=0x0, argcount=0, kws=0x0, kwcount=0, defs=0x0, defcount=0, closure=0x0) at Python/ceval.c:2831 #15 0x080c6647 in PyEval_EvalCode (co=0xb7e6c608, globals=0xb7ebeacc, locals=0xb7ebeacc) at Python/ceval.c:494 #16 0x080e52d8 in PyRun_FileExFlags (fp=0x8e8e488, filename=0xbff8cb6f /home/jaap/downloads/sage-2.9.2/local/bin/sage-gdb-pythonstartup, ---Type return to continue, or q return to quit--- start=257, globals=0xb7ebeacc, locals=0xb7ebeacc, closeit=0, flags=0xbff8bdb8) at Python/pythonrun.c:1271 #17 0x080e5567 in PyRun_SimpleFileExFlags
[sage-devel] Re: SAGE 2.10.alpha0 released
On Jan 8, 5:04 pm, Kiran Kedlaya [EMAIL PROTECTED] wrote: Hi Kiran, Works for me. Good to know, but that was the intention all the way. That means #1553 is solved, though the larger SCons issue remains in case we bring in any more software that uses it. I meant to write the PoLyBoRi team an email with the patch so they can integrate it upstream. I see no reason not to import PATH from ENV. On my box (again, that's 64-bit RHEL5), sage -testall passes all tests. Excellent. Kiran Cheers, Michael --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~--~~~~--~~--~--~---
[sage-devel] Re: Anyone knows about any regression in gmp-4.2.2?
On Jan 8, 11:55 pm, mabshoff [EMAIL PROTECTED]
dortmund.de wrote:
On Jan 8, 10:00 am, Francois [EMAIL PROTECTED] wrote:
Hi,
Hi Francois,
In my efforts to get sage in Gentoo I came on something that looks
like a problem in gmp 4.2.2.
On Gentoo gmp 4.2.2 is marked stable and is part of the system (needed
by gcc). On the ground
that the only patch relevant to me was the new fast gcd code and that
it was a performance patch
and not a correctness one, I decided to use my system gmp (along with
bzip2, readline, mpfr,gd,
pari [mistake since the Galois data are not included in Gentoo],
gnutls, atlas [oh yes not another
5 hours of tunning, please] and maxima - I passed on python for now).
Having a successful build after rather minimal adjustment I decided to
test it
(sage -testall) and compare the results with a regular build. Numerous
failures at
various point. First tut.tex:
sage -t tut.tex
***
***
File tut.py, line 1126:
: x = crt(2, 1, 3, 5); x
Expected:
11
Got:
-4
**
File tut.py, line 2250:
: M.T(11).charpoly('x').factor()
Expected:
(x - 285311670612) * (x - 534612)^2
Got:
x^3 - 732255212432452092*x^2 + 732255211931384496*x -
3145012477679296873951599424
**
2 items had failures:
1 of 8 in __main__.example_48
1 of 8 in __main__.example_96
***Test Failed*** 2 failures.
For whitespace errors, see the file .doctest_tut.tex
[60.2 s]
exit code: 256
and then a few more. For the first failed test putting back gmp/mpfr
in the build
solved the problem for the second putting back pari worked [of course
system
pari was built against system gmp so it probably propagated from
there]. That
solved almost all the failed test I had (the test didn't finish got
stuck in calc.py
if memory serves me correctly).
The system's Maxima ought to be at fault here.
I checked that gmp was built properly with the make check provided
in gmp,
so it doesn't look like a miscompilation.
You may have hithttp://trac.sagemath.org/sage_trac/ticket/1482since
the system gmp didn't have the improved gcd code patched in. The
default behavior is not to pass the parameter minimal, so the issues
you see won't be fixed if you use a non-patched gmp. For many people
patching the gmp with a GPLed patch making the derived work GPL only
will be a problem since many distributions ship code that depends on
GPL being LGPL.
I guess it should be investigated before this version of gmp makes it
into sage
I am confident it will be resolved. The changes gmp 4.2.1-4.2.2 were
largely config fixes and fixes for exotic platforms, so I don't expect
any problems once we remerge our patch set.
I didn't know about that minimal parameter - I made the assumption
that
the interfacing wouldn't be changed for a minor version, I can
understand
as John Cremona that in fact the test I show is correct. I am more
shocked
at the behavior of pari I don't know if it is to be expected with an
unpatched
gmp. I may try to patch my system gmp and see what happens with that.
I am sure that system maxima is at fault. I just like to know where it
originates
from. mpfr? (I realised recently fentoo use 2.3.0.p3 not p4 - it is in
portage but in
unstable probably no one care enough to stabilize it fast) lisp? The
default
lisp for maxima here is the Steel Bank Common Lisp and is not
callable
from sage, I got hit by http://trac.sagemath.org/sage_trac/ticket/1002
(or http://bugs.gentoo.org/show_bug.cgi?id=54738 ), gcl fails I had to
use
CMU common lisp. Something else?
Did you open http://trac.sagemath.org/sage_trac/ticket/1721
because of my remark on atlas? It has to come with a warning to set
SAGE_FORTRAN as well.
--~--~-~--~~~---~--~~
To post to this group, send email to sage-devel@googlegroups.com
To unsubscribe from this group, send email to [EMAIL PROTECTED]
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/
-~--~~~~--~~--~--~---
[sage-devel] Re: Anyone knows about any regression in gmp-4.2.2?
On Jan 8, 7:40 pm, Francois [EMAIL PROTECTED] wrote: SNIP I am confident it will be resolved. The changes gmp 4.2.1-4.2.2 were largely config fixes and fixes for exotic platforms, so I don't expect any problems once we remerge our patch set. I didn't know about that minimal parameter - I made the assumption that the interfacing wouldn't be changed for a minor version, Just to be clear: That is a parameter on our end to guarantee minimality for people who demand it. When David Harvey added Nils Möller's extended gcd patch to our gmp.spkg William and I ended up fixing all the doctests that changed due to the slightly different, but still mathematically correct results. I can understand as John Cremona that in fact the test I show is correct. I am more shocked at the behavior of pari I don't know if it is to be expected with an unpatched gmp. I may try to patch my system gmp and see what happens with that. Ok, I am lost what could have caused issues with pari since I don't know what Gentoo does regarding patches, version and so oin. I am sure that system maxima is at fault. I just like to know where it originates from. mpfr? (I realised recently fentoo use 2.3.0.p3 not p4 - it is in portage but in unstable probably no one care enough to stabilize it fast) I doubt that is the issue. When I upgraded mpfr from 2.3.0 to 2.3.0.p4 none of the doctests were affected, so 2.3.0.p3 should be fine. You should be aware that we will upgrade to mpfr 2.3.1 once it is out in a couple days. lisp? The default lisp for maxima here is the Steel Bank Common Lisp and is not callable from sage, I got hit byhttp://trac.sagemath.org/sage_trac/ticket/1002 (orhttp://bugs.gentoo.org/show_bug.cgi?id=54738), gcl fails I had to use CMU common lisp. Something else? I don't think the behavior of Maxima is affected by lisp. What version of Maxima do you run on your system? There is at least one bug in the Sage-Maxima interface that we are aware of, but that one is triggered on Arch Linux. What you experience is exactly the reason Sage is a more or less complete stand alone source tarball. You only replaced a couple packages and are already knee deep in a mess that is hard to debug. Now imagine we had to do that using all distribution packages and doing that for several distributions. The integration with Debian is only possible because we will either maintain the packages ourselves or closely coordinate with the current maintainers. Any other way will be very difficult. Did you openhttp://trac.sagemath.org/sage_trac/ticket/1721 because of my remark on atlas? It has to come with a warning to set SAGE_FORTRAN as well. Well, I had thought about it for a while, ever since reports about failed builds and long builds in general popped up. I can certainly understand why this is convenient for some people and the variety of ATLAS releases is small. It wouldn't be the default, but for those who know what they do it would be a convenient feature. It is also geared toward people who download binaries but do not have a current enough CPU, i.e. SSE2. So they could just install ATLAS from the distribution and be done with it. Cheers, Michael --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~--~~~~--~~--~--~---
[sage-devel] Re: Twisted authentication issue
On Jan 8, 2008 1:46 AM, Robert Bradshaw [EMAIL PROTECTED] wrote: Hey Alex, Hi Robert, I'm stumped by some of the twisted notebook code. Specifically, if the login form is set to POST instead of GET it redirects login/ You are definitely going to want the login form to always be a POST not a GET, so that the credentials are sent as post-data, not as GET url parameters. This slash is particularly annoying. In what case is this causing a problem? Is there any way to 1) have it not redirect to login/ the path 'login' is required to be there for the authentication to work (you could change it to some other name, but then you would have to change other code) or 2) set the session cookie path (to '/') You can always set the cookie path to whatever you want when you initially send it (right after login) through the 'set-cookie' http header. I bet you have similar issues knoboo. I have definitely made some improvements in Knoboo to the authentication code since it was first put into the Sage notebook. One major issue that still exists in the Sage notebook has to do with login and is *a security issue*: It is necessary to do a redirect after 'login' (which is a POST), to a GET such that the post-data is not stored in the browser. This is an important issue because *after logout the username/password remains in the browser* and all someone else has to do to login with the last person's credentials is to hit the back-button and then refresh the page. (Try it yourself) The solution to this problem is called the 'post-redirect-get method' (see google for more info) and besides be needed for security, I think it might solve the above problems that you are facing. Alex Thanks, - Robert --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~--~~~~--~~--~--~---
[sage-devel] Re: SAGE 2.10.alpha0 released
On Tue, 8 Jan 2008 09:34:06 -0800 (PST) mabshoff [EMAIL PROTECTED] wrote: On Jan 8, 5:04 pm, Kiran Kedlaya [EMAIL PROTECTED] wrote: That means #1553 is solved, though the larger SCons issue remains in case we bring in any more software that uses it. I meant to write the PoLyBoRi team an email with the patch so they can integrate it upstream. I see no reason not to import PATH from ENV. They already know about it, and they will incorporate the fix. Maybe we should consider patching the SCons package, so that it uses better default values. I thought this was going to be the solution for #1656. Burcin --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~--~~~~--~~--~--~---
[sage-devel] Re: Anyone knows about any regression in gmp-4.2.2?
On Jan 9, 8:14 am, mabshoff [EMAIL PROTECTED] dortmund.de wrote: On Jan 8, 7:40 pm, Francois [EMAIL PROTECTED] wrote: SNIP I don't think the behavior of Maxima is affected by lisp. What version of Maxima do you run on your system? There is at least one bug in the Sage-Maxima interface that we are aware of, but that one is triggered on Arch Linux. maxima-5.13, upgrading to 5.14 in a few minutes. The only obvious patch in Gentoo is about xdg-utils. What you experience is exactly the reason Sage is a more or less complete stand alone source tarball. You only replaced a couple packages and are already knee deep in a mess that is hard to debug. Now imagine we had to do that using all distribution packages and doing that for several distributions. The integration with Debian is only possible because we will either maintain the packages ourselves or closely coordinate with the current maintainers. Any other way will be very difficult. I wasn't expecting it to be easy. In fact after some thinking, it could only live in an overlay as the release are coming to fast for it to make it into stable (one month without open bugs - it could never be current). I was careful in my choice of packages that the patching was minimal/ doc to non-existant. Which is why I chickened on anything that was python related. So yeah I am still surprised at the amount of poo I gathered. I still think the experience is/was worthwhile but I am not sure I have the resources to go for it alone - I know there are a few people interested in Gentooland. I certainly would need to seriously upgrade my home machine. Did you openhttp://trac.sagemath.org/sage_trac/ticket/1721 because of my remark on atlas? It has to come with a warning to set SAGE_FORTRAN as well. Well, I had thought about it for a while, ever since reports about failed builds and long builds in general popped up. I can certainly understand why this is convenient for some people and the variety of ATLAS releases is small. It wouldn't be the default, but for those who know what they do it would be a convenient feature. It is also geared toward people who download binaries but do not have a current enough CPU, i.e. SSE2. So they could just install ATLAS from the distribution and be done with it. I can see where that come from. Compile time was relatively short with 3.7.x as there was a profile for my machine. 3.8.0 was a shocker, so sage asking me to compile it again, for its internal purpose, was certainly not welcome. Which kinds of bring to another thing that I wanted to ask about for a while. I noticed that there is in fact 2 blas packages in sage, atlas and what is usually refereed as blas-reference in Gentoo. It seems that the goal of that one is just to provide f77blas. Do we *really* need that? --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~--~~~~--~~--~--~---
[sage-devel] Re: Anyone knows about any regression in gmp-4.2.2?
On Jan 8, 8:48 pm, Francois [EMAIL PROTECTED] wrote: On Jan 9, 8:14 am, mabshoff [EMAIL PROTECTED] dortmund.de wrote: On Jan 8, 7:40 pm, Francois [EMAIL PROTECTED] wrote: SNIP I don't think the behavior of Maxima is affected by lisp. What version of Maxima do you run on your system? There is at least one bug in the Sage-Maxima interface that we are aware of, but that one is triggered on Arch Linux. maxima-5.13, upgrading to 5.14 in a few minutes. The only obvious patch in Gentoo is about xdg-utils. What you experience is exactly the reason Sage is a more or less complete stand alone source tarball. You only replaced a couple packages and are already knee deep in a mess that is hard to debug. Now imagine we had to do that using all distribution packages and doing that for several distributions. The integration with Debian is only possible because we will either maintain the packages ourselves or closely coordinate with the current maintainers. Any other way will be very difficult. I wasn't expecting it to be easy. In fact after some thinking, it could only live in an overlay as the release are coming to fast for it to make it into stable (one month without open bugs - it could never be current). I was careful in my choice of packages that the patching was minimal/ doc to non-existant. Which is why I chickened on anything that was python related. So yeah I am still surprised at the amount of poo I gathered. I still think the experience is/was worthwhile but I am not sure I have the resources to go for it alone - I know there are a few people interested in Gentooland. I certainly would need to seriously upgrade my home machine. Ok, we would certainly welcome any kind of participation from the Gentoo community. Did you openhttp://trac.sagemath.org/sage_trac/ticket/1721 because of my remark on atlas? It has to come with a warning to set SAGE_FORTRAN as well. Well, I had thought about it for a while, ever since reports about failed builds and long builds in general popped up. I can certainly understand why this is convenient for some people and the variety of ATLAS releases is small. It wouldn't be the default, but for those who know what they do it would be a convenient feature. It is also geared toward people who download binaries but do not have a current enough CPU, i.e. SSE2. So they could just install ATLAS from the distribution and be done with it. I can see where that come from. Compile time was relatively short with 3.7.x as there was a profile for my machine. 3.8.0 was a shocker, so sage asking me to compile it again, for its internal purpose, was certainly not welcome. Which kinds of bring to another thing that I wanted to ask about for a while. I noticed that there is in fact 2 blas packages in sage, atlas and what is usually refereed as blas-reference in Gentoo. It seems that the goal of that one is just to provide f77blas. Do we *really* need that? It depends. We need to build Lapack before ATLAS since we build an improved liblapack with the Lapack bits from ATLAS. Now if we were to run Lapack's test suite we need a BLAS. Atlas is also not standard yet on OSX, so there are more considerations here. Since netlib.org's BLAS is small and compiles quickly nobody has advocated its removal and since I consider it important to be able to run the Lapack testers on demand I would like it to stay. Cheers, Michael --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~--~~~~--~~--~--~---
[sage-devel] SCREMS -- feedback needed
Hi, Robert Miller and I are trying hard to get some sexy new hardware via the NSF SCREMS program. Please see and comment on the attached proposal. Any typos, thoughts, etc., are welcome. http://sage.math.washington.edu/grants/screms/ NSF evaluates proposals based on scientific value, etc., and this proposal is mainly about number theory and other research because of that. But the equipment would also greatly benefit the Sage project. This proposal will be submitted on Thursday so please make comments asap. -- William -- William Stein Associate Professor of Mathematics University of Washington http://wstein.org --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~--~~~~--~~--~--~---
[sage-devel] Re: Wiris -- something like the Sage notebook sort of
On Jan 7, 2008, at 23:50 , William Stein wrote: On Jan 7, 2008 11:42 PM, Francois [EMAIL PROTECTED] wrote: On Jan 8, 8:22 pm, Fernando Perez [EMAIL PROTECTED] wrote: On Jan 8, 2008 12:13 AM, Francois [EMAIL PROTECTED] wrote: FWIW, it worked for me under Ubuntu Gutsy with firefox... Strange it definitely just stay blank for me. I am using the plugin from the sun jdk 1.6, do you use 1.5 or 1.6 in Ubuntu? I'm using OSX 10.5.1 on a standard mac laptop, and I just get a blank screen too. That's weird. On my Core Duo (10.5.1), it works just fine...(I tried the CAS, and clicked the represent - plotter line). Is there anything showing up in your logs when you do this? I usually open Console, select all messages and then insert marker before trying something I'm tracking down. That helps keep the chaos from the logs manageable. Justin -- Justin C. Walker, Curmudgeon-at-Large () The ASCII Ribbon Campaign /\ Help Cure HTML Email --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~--~~~~--~~--~--~---
[sage-devel] Re: Wiris -- something like the Sage notebook sort of
On Jan 8, 2008 12:07 PM, Justin C. Walker [EMAIL PROTECTED] wrote: On Jan 7, 2008, at 23:50 , William Stein wrote: On Jan 7, 2008 11:42 PM, Francois [EMAIL PROTECTED] wrote: On Jan 8, 8:22 pm, Fernando Perez [EMAIL PROTECTED] wrote: On Jan 8, 2008 12:13 AM, Francois [EMAIL PROTECTED] wrote: FWIW, it worked for me under Ubuntu Gutsy with firefox... Strange it definitely just stay blank for me. I am using the plugin from the sun jdk 1.6, do you use 1.5 or 1.6 in Ubuntu? I'm using OSX 10.5.1 on a standard mac laptop, and I just get a blank screen too. That's weird. On my Core Duo (10.5.1), it works just fine...(I tried the CAS, and clicked the represent - plotter line). Is there anything showing up in your logs when you do this? I usually open Console, select all messages and then insert marker before trying something I'm tracking down. That helps keep the chaos from the logs manageable. It works for me now. I think there was a problem because the network I was on was slow /flakie. -- William --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~--~~~~--~~--~--~---
[sage-devel] Re: Twisted authentication issue
On Jan 8, 2008, at 11:09 AM, alex clemesha wrote: On Jan 8, 2008 1:46 AM, Robert Bradshaw [EMAIL PROTECTED] wrote: Hey Alex, Hi Robert, I'm stumped by some of the twisted notebook code. Specifically, if the login form is set to POST instead of GET it redirects login/ You are definitely going to want the login form to always be a POST not a GET, so that the credentials are sent as post-data, not as GET url parameters. Yes, the login form should always be POST, but here I do want to use GET parameters. Specifically, when the server starts up, it can provide a URL with a one-time authentication token in the URL it pops up. Since, in our case, users are often starting up their own servers this will be really handy (and not a security risk, because they could just do reset=True anyways. This slash is particularly annoying. In what case is this causing a problem? Is there any way to 1) have it not redirect to login/ the path 'login' is required to be there for the authentication to work (you could change it to some other name, but then you would have to change other code) login is fine, login redirecting to login/ is causing the error. or 2) set the session cookie path (to '/') You can always set the cookie path to whatever you want when you initially send it (right after login) through the 'set-cookie' http header. I wasn't able to figure out where, exactly, set-cookie was getting called. It seemed to be called somewhere in the bowels of the twisted cred module, and I didn't know how to access it. This would be perfect, if it works. I bet you have similar issues knoboo. I have definitely made some improvements in Knoboo to the authentication code since it was first put into the Sage notebook. One major issue that still exists in the Sage notebook has to do with login and is *a security issue*: It is necessary to do a redirect after 'login' (which is a POST), to a GET such that the post-data is not stored in the browser. This is an important issue because *after logout the username/password remains in the browser* and all someone else has to do to login with the last person's credentials is to hit the back-button and then refresh the page. (Try it yourself) The solution to this problem is called the 'post-redirect-get method' (see google for more info) and besides be needed for security, I think it might solve the above problems that you are facing. Thanks for your response, and, we should look into this issue too. - Robert --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~--~~~~--~~--~--~---
[sage-devel] Re: SAGE 2.10.alpha0 released
On Jan 8, 8:54 pm, Burcin Erocal [EMAIL PROTECTED] wrote: On Tue, 8 Jan 2008 09:34:06 -0800 (PST) mabshoff [EMAIL PROTECTED] wrote: On Jan 8, 5:04 pm, Kiran Kedlaya [EMAIL PROTECTED] wrote: That means #1553 is solved, though the larger SCons issue remains in case we bring in any more software that uses it. I meant to write the PoLyBoRi team an email with the patch so they can integrate it upstream. I see no reason not to import PATH from ENV. Hi Burcin, They already know about it, and they will incorporate the fix. Ok, I send the patch to Michael B. today off list, but I hadn't heard back yet. I also think that that fix is a potential solution for the Itanium issue that William reported since at least on the box I am currently playing with a gcc 3.2 gets picked and that doesn't end too well as you can imagine. I plan to build a new gcc 4.2.2 on that box tonight and also build 2.10.alpha0 on there to see if the problem is resolved. Maybe we should consider patching the SCons package, so that it uses better default values. I thought this was going to be the solution for #1656. Well, #1656 is resolved since we took care of both spkgs that use SCons. I don't think there are any better default values, especially since choice A that works on one system will surely not work on another. We just need to make sure to check any new code that uses SCons also imports PATH from ENV. Burcin Cheers, Michael --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~--~~~~--~~--~--~---
[sage-devel] Re: Anyone knows about any regression in gmp-4.2.2?
On Jan 9, 8:58 am, mabshoff [EMAIL PROTECTED] dortmund.de wrote: Ok, we would certainly welcome any kind of participation from the Gentoo community. Hopefully some support may be drummed there, I myself may be soon claimed by a non-academic real-life that could put me on hiatus. Which kinds of bring to another thing that I wanted to ask about for a while. I noticed that there is in fact 2 blas packages in sage, atlas and what is usually refereed as blas-reference in Gentoo. It seems that the goal of that one is just to provide f77blas. Do we *really* need that? It depends. We need to build Lapack before ATLAS since we build an improved liblapack with the Lapack bits from ATLAS. Now if we were to run Lapack's test suite we need a BLAS. Atlas is also not standard yet on OSX, so there are more considerations here. Since netlib.org's BLAS is small and compiles quickly nobody has advocated its removal and since I consider it important to be able to run the Lapack testers on demand I would like it to stay. Ok make sense, it's one thing that I had disabled without much side effect apart from touching cvxopt about lf77blas although I believe numpy, scipy and al will detect and use it if present. Plus while building with my system atlas it failed because it wanted to remove /usr/lib/libblas.so. Something to think about for your SYSTEM_ATLAS setting. --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~--~~~~--~~--~--~---
[sage-devel] Re: SCREMS -- feedback needed
Good luck! @ Isn't it a bit obvious to say that you will be doing computations which go beyond what has been done before? @ You say that the previous NSF grant has had a broad impact on mathematics -- that should be justified or it just sounds like hype. [Later I see that you do justify this below, so all that's needed is a cross-reference here.] @ On the undergraduate's role I would emphasis what they will learn mathematically, and have the sytems admin role secondary. Whatever it will actually be like in practice, there will be people at NSF who regrd systems admin as menial... @ You describe your recent joint work with Mazur as unpublished. Perhaps better to say not yet published? @ You'll need to explain the web services you will be running. They might not like it if they think that the machines could be used by anyone in the world, for not necessarily deserving work! @ I concur with Josh Cantor's comments (despite the grinding to a halt bit ;)) Which I later see is a quote! Fair though. @ Include Compute all Q-curves ... ? Thanks to Serre's conjecture we now know that these are all the modular elliptic curves over number fields. @ Reference [Cre97] needs updating (nott - warwick) I hope they ask me to referee this one! John On 08/01/2008, William Stein [EMAIL PROTECTED] wrote: Hi, Robert Miller and I are trying hard to get some sexy new hardware via the NSF SCREMS program. Please see and comment on the attached proposal. Any typos, thoughts, etc., are welcome. http://sage.math.washington.edu/grants/screms/ NSF evaluates proposals based on scientific value, etc., and this proposal is mainly about number theory and other research because of that. But the equipment would also greatly benefit the Sage project. This proposal will be submitted on Thursday so please make comments asap. -- William -- William Stein Associate Professor of Mathematics University of Washington http://wstein.org -- John Cremona --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~--~~~~--~~--~--~---
[sage-devel] Re: Anyone knows about any regression in gmp-4.2.2?
On Jan 8, 9:55 pm, Francois [EMAIL PROTECTED] wrote: On Jan 9, 8:58 am, mabshoff [EMAIL PROTECTED] dortmund.de wrote: Ok, we would certainly welcome any kind of participation from the Gentoo community. Hopefully some support may be drummed there, Excellent. I myself may be soon claimed by a non-academic real-life that could put me on hiatus. Sooner or later it happens to most of us ;) Which kinds of bring to another thing that I wanted to ask about for a while. I noticed that there is in fact 2 blas packages in sage, atlas and what is usually refereed as blas-reference in Gentoo. It seems that the goal of that one is just to provide f77blas. Do we *really* need that? It depends. We need to build Lapack before ATLAS since we build an improved liblapack with the Lapack bits from ATLAS. Now if we were to run Lapack's test suite we need a BLAS. Atlas is also not standard yet on OSX, so there are more considerations here. Since netlib.org's BLAS is small and compiles quickly nobody has advocated its removal and since I consider it important to be able to run the Lapack testers on demand I would like it to stay. Ok make sense, it's one thing that I had disabled without much side effect apart from touching cvxopt about lf77blas although I believe numpy, scipy and al will detect and use it if present. Plus while building with my system atlas it failed because it wanted to remove /usr/lib/libblas.so. Something to think about for your SYSTEM_ATLAS setting. Ok, that does seem an odd thing to do. Josh told me a while back that numpy, scipy and cvxopt all use atlas per default, so maybe there is something different the way ATLAS is shipped on Gentoo. Sooner or later I will remove F77 BLAS and see what happens ;) Cheers, Michael --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~--~~~~--~~--~--~---
[sage-devel] Re: SCREMS -- feedback needed
Sorry I didn't read it all - a bit busy tonight, but some comments: There is a typo on the third last line of the summary: research - researchers ? The phrase impact on mathematics or something equivalent is used too often in the summary. It's too generic. I personally think it is a good move mentioning the Sage supercomputer. Personally I'd mention it earlier on and mention the press it got, including overseas. I think this is an extremely important point, since that machine has opened up a whole world of possibilities. The E8 computation alone has spawned an immense amount of interest, let alone all the other stuff done on there. I'm confused about the hardware itself. You mention two 16 core servers in the main description, but in the budget justification (I may have missed them) I don't see them. Instead I see 6 high end desktop computers. Perhaps this is something to do with what the grant guidelines tell you to do. I haven't read those. John already picked up numerous things which I noted, (so I won't duplicate them), and I strongly agree with his comments. Bill. William Stein wrote: Hi, Robert Miller and I are trying hard to get some sexy new hardware via the NSF SCREMS program. Please see and comment on the attached proposal. Any typos, thoughts, etc., are welcome. http://sage.math.washington.edu/grants/screms/ NSF evaluates proposals based on scientific value, etc., and this proposal is mainly about number theory and other research because of that. But the equipment would also greatly benefit the Sage project. This proposal will be submitted on Thursday so please make comments asap. -- William -- William Stein Associate Professor of Mathematics University of Washington http://wstein.org --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~--~~~~--~~--~--~---
[sage-devel] Education: Tying SAGE into an online course management system
For people who are seriously interested into developing SAGE into a tool that can easily be used for teaching and/or student assignments, I warmly recommend that they look into LON-CAPA. It has the advantage that it is already based on open source technology. Maxima is already packaged with it for symbolic capabilities. One simple integration I would imagine is tying SAGE into LON-CAPA's authentication system (or at least provide a plug-in to that effect). Presently, one big obstacle to offering SAGE as a tool for a largish class is the account management (together with the problems of running sage as a widely accessible service in a secure way) One could just say use sagenb.org in class, but I imagine that could end the offering of that service. There is going to be a LON-CAPA conference in May at SFU: http://www.sfu.ca/~capachem/conference/ There are some lecturers at SFU who are quite knowledgeable concerning LON-CAPA and may well be interested in seeing how SAGE and LON-CAPA could be integrated to advantage of both. --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~--~~~~--~~--~--~---
[sage-devel] Re: Wiris -- something like the Sage notebook sort of
Francois wrote: And it is supposed to work with what? On Linux I get a blank page with firefox and konqueror - opera just went and crashed. I certainly won't take seriously a product of that kind, that I cannot test on Linux. It takes a while for the applet to load the first time. I had to wait for a minute or two for it to fully load. Ted --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~--~~~~--~~--~--~---
[sage-devel] Re: Wiris -- something like the Sage notebook sort of
Justin C. Walker wrote: A brief comment: I think Harald's comment is apt. +1 Jaap --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~--~~~~--~~--~--~---
[sage-devel] Re: Wiris -- something like the Sage notebook sort of
Ted Kosan wrote: Harald wrote: but I'm more and more convinced that a huge java applet is not the best way for any webbased project. There are two things to consider: 1. it is much heavier, concerning load time/memory and so on. With bandwidth and memory capacity increasing exponentially, this is fast becoming a non-issue. I've heard that before! But I don't believe this is the panacea! it depends on installing java, that's not default on win32! But Java installs so easily on win32 that I don't know anyone who has a problem doing this anymore. Beyond this, Java runs reasonably uniformly across the major browsers which will avoid most of the multi-browser difficulties the Javascript notebook faces. Another plus is that most cell phones also run Java. -1 and 2. on the javascript side, in the future a lot will change when there are good JIT compilers. this could make js-code possibly ~10-100x faster! Javascript will become quicker but what I am even more interested in are Java's huge libraries and the huge and growing amount of open source software that is written in Java. I don't see Javascript matching this any time soon, if ever. Beyond this, Java is still the most popular programming language in existence (http://www.tiobe.com/tpci.htm) and so it is an extremely safe bet by almost any metric :-) -1 Maybe java is popular, but ask your self why? Managers want it, but they usually don't know why! Ok, I once thought java was the way to go, but now I'm a convert. I hate java. E.g. why the hell the new jmoll graphics does not work on my Fedora 7 install from the sage command line? Jaap --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~--~~~~--~~--~--~---
[sage-devel] Re: Wiris -- something like the Sage notebook sort of
Harald wrote: but I'm more and more convinced that a huge java applet is not the best way for any webbased project. There are two things to consider: 1. it is much heavier, concerning load time/memory and so on. With bandwidth and memory capacity increasing exponentially, this is fast becoming a non-issue. it depends on installing java, that's not default on win32! But Java installs so easily on win32 that I don't know anyone who has a problem doing this anymore. Beyond this, Java runs reasonably uniformly across the major browsers which will avoid most of the multi-browser difficulties the Javascript notebook faces. Another plus is that most cell phones also run Java. and 2. on the javascript side, in the future a lot will change when there are good JIT compilers. this could make js-code possibly ~10-100x faster! Javascript will become quicker but what I am even more interested in are Java's huge libraries and the huge and growing amount of open source software that is written in Java. I don't see Javascript matching this any time soon, if ever. Beyond this, Java is still the most popular programming language in existence (http://www.tiobe.com/tpci.htm) and so it is an extremely safe bet by almost any metric :-) Ted --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~--~~~~--~~--~--~---
[sage-devel] Re: Wiris -- something like the Sage notebook sort of
A brief comment: On Jan 8, 2008, at 2:29 PM, Ted Kosan wrote: Harald wrote: but I'm more and more convinced that a huge java applet is not the best way for any webbased project. There are two things to consider: 1. it is much heavier, concerning load time/memory and so on. With bandwidth and memory capacity increasing exponentially, this is fast becoming a non-issue. I disagree with this. In the same way that the stock market, say, builds in the common wisdom about the future into the current price of any stock being traded, developers are building in an expectation of future bandwidth and capacity to the products they are developing now. The overall performance gain for new hardware does not compare to the actual raw numbers, since the software is generally at or ahead of that performance curve. Is running Vista on last year's ThinkTank recommended? I think Harald's comment is apt. Justin -- Justin C. Walker, Curmudgeon-At-Large, Director Institute for the Enhancement of the Director's Income The path of least resistance: it's not just for electricity any more. --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~--~~~~--~~--~--~---
[sage-devel] Fwd: these are the links I mentioned...
Sage devel might be interested in this -- it's screencasts of a tool that numerical python people that DoD / supercomputing sorts have developed -- Forwarded message -- From: Jose Unpingco Date: Jan 8, 2008 11:29 AM Subject: these are the links I mentioned... To: [EMAIL PROTECTED] Nice meeting you. Keep up the great work with SAGE. The two following links might interest you as we are currently using VISION as a visually programmable front-end for parallel computing at SSC-SD as shown in the following: www.osc.edu/~unpingco/Tutorial_11Dec.html http://www.osc.edu/~unpingco/Tutorial_03Jan2008.html Please contact me if you have questions or need more information. Thanks! Jose Unpingco, Ph.D. -- William Stein Associate Professor of Mathematics University of Washington http://wstein.org --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~--~~~~--~~--~--~---
[sage-devel] Re: Education: Tying SAGE into an online course management system
On Jan 9, 12:42 am, mhampton [EMAIL PROTECTED] wrote: I encourage more thought and discussion on this. Yes, this should be discussed. I think the shortest path at first would be integrating SAGE-evaluation-boxes in the already shipped moinmoin Wiki. This is then a nice place to build a local knowledge base. And account management is also very important (also LDAP and so on. maybe there are already python based universal login modules? I don't know) -- and some sort of group management and permissions would be also nice. But I think the first step is to improve quality and polish the system. h --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~--~~~~--~~--~--~---
[sage-devel] Re: Education: Tying SAGE into an online course management system
Here at the joint meetings, there has been some interest from some course management folks in working with sage. In particular, someone from WebWork stopped by and suggested we should look for complementary features. I think some sort of open source course management functionality in sage would be great; it would add a lot of bulk to actually incorporate into sage, as opposed to mutually design for interoperability (a sort of plugin approach). I encourage more thought and discussion on this. cheers, Marshall Hampton On Jan 8, 1:16 pm, Nils Bruin [EMAIL PROTECTED] wrote: For people who are seriously interested into developing SAGE into a tool that can easily be used for teaching and/or student assignments, I warmly recommend that they look into LON-CAPA. It has the advantage that it is already based on open source technology. Maxima is already packaged with it for symbolic capabilities. One simple integration I would imagine is tying SAGE into LON-CAPA's authentication system (or at least provide a plug-in to that effect). Presently, one big obstacle to offering SAGE as a tool for a largish class is the account management (together with the problems of running sage as a widely accessible service in a secure way) One could just say use sagenb.org in class, but I imagine that could end the offering of that service. There is going to be a LON-CAPA conference in May at SFU: http://www.sfu.ca/~capachem/conference/ There are some lecturers at SFU who are quite knowledgeable concerning LON-CAPA and may well be interested in seeing how SAGE and LON-CAPA could be integrated to advantage of both. --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~--~~~~--~~--~--~---
[sage-devel] Re: Fwd: these are the links I mentioned...
On Jan 8, 2008 4:18 PM, William Stein [EMAIL PROTECTED] wrote: Sage devel might be interested in this -- it's screencasts of a tool that numerical python people that DoD / supercomputing sorts have developed Wow, this is cool. We have an ipython1 screencast now :) I'd never heard of this... Vision is really neat, BTW, Developed at the Scripps institute by a very good team led by Michel Sanner. They've presented at various scipy conferences: http://www.scripps.edu/~sanner/ Cheers, f --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~--~~~~--~~--~--~---
[sage-devel] Re: Wiris -- something like the Sage notebook sort of
Jaap wrote: With bandwidth and memory capacity increasing exponentially, this is fast becoming a non-issue. I've heard that before! But I don't believe this is the panacea! The exponential growth of computing indicates that by 2020, a $1000 computer will have the same capacity as a human brain. By 2050, a $1000 computer will have the same capacity as all the human brains on the planet. http://en.wikipedia.org/wiki/Image:PPTExponentialGrowthof_Computing.jpg http://www.kurzweilai.net/meme/frame.html?main=/articles/art0134.html? To me, exponential growth seems more like magic than anything else, even if it is just simple science :-) If these projections are even close to being accurate then computer capacity is not going to be a problem for Java. I truly believe that the real issue we will be dealing with in the not so distant future is the best way to upload Sage into human brains. Maybe java is popular, but ask your self why? Managers want it, but they usually don't know why! Java is one of the most popular languages for manager-free open source projects on sourceforge too: http://www.cs.berkeley.edu/~flab/languages.html I agree, though, that Java has a significant amount of hype associated with it and this is definitely a distraction. Ok, I once thought java was the way to go, but now I'm a convert. I hate java. Truth be told, I like programming in Python better than Java when this is feasible. For rich cross-platform browser-based applications, however, it is difficult to beat. E.g. why the hell the new jmoll graphics does not work on my Fedora 7 install from the sage command line? Some of this might be due to the quickness with which jmol was added to Sage so that it could be demonstrated at the AMS meeting. Ted --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~--~~~~--~~--~--~---
[sage-devel] Re: Wiris -- something like the Sage notebook sort of
On Jan 9, 1:31 am, Ted Kosan [EMAIL PROTECTED] wrote: Jaap wrote: With bandwidth and memory capacity increasing exponentially, this is fast becoming a non-issue. I've heard that before! But I don't believe this is the panacea! The exponential growth of computing indicates that by 2020, a $1000 computer will have the same capacity as a human brain. By 2050, a $1000 computer will have the same capacity as all the human brains on the planet. http://en.wikipedia.org/wiki/Image:PPTExponentialGrowthof_Computing.jpg http://www.kurzweilai.net/meme/frame.html?main=/articles/art0134.html? To me, exponential growth seems more like magic than anything else, even if it is just simple science :-) Well, in the current form it is doubtful that Moore's law will hold that long, but the expectation that Moore's law will take care of any kind of performance issue by just providing enough power gave us Windows Vista. This is obviously a rim shot, but just because it is there isn't a valid argument. I started my computing career with 16 kb of RAM - it has been a while. And while I certainly don't remember the time of 40x25 16 color displays one has to wonder every once in a while where things could have gone better in OS design in the last 20 years. If these projections are even close to being accurate then computer capacity is not going to be a problem for Java. I truly believe that the real issue we will be dealing with in the not so distant future is the best way to upload Sage into human brains. Maybe java is popular, but ask your self why? Managers want it, but they usually don't know why! Java is one of the most popular languages for manager-free open source projects on sourceforge too: http://www.cs.berkeley.edu/~flab/languages.html Well, that doesn't really mean much - the interesting question is how many of those projects are thriving and actually used by people. Azureus is the only Java app I have ever looked at and was willing to use. I switched away from it because I removed Java from my box due to the constant need to update it for security reasons. Especially on Windows you have to remove older JDKs manually since they stick around and since the java applet can request specific JDK versions, even older installed ones, that is a big potential infection vector. I agree, though, that Java has a significant amount of hype associated with it and this is definitely a distraction. Ok, I once thought java was the way to go, but now I'm a convert. I hate java. Truth be told, I like programming in Python better than Java when this is feasible. For rich cross-platform browser-based applications, however, it is difficult to beat. Well, some people will tell you that running real application in a browser is the problem in the first place, but I don't intent to start a flame war. Java has its place, but I personally dislike the GUI interface interface and I don't see the need for command line applets. We have that think called C that is cross platform :). Even eclipse based code is slow and every time somebody tries to convince me to use an IDE which requires a couple hundred MB to run, even more space to install and feels like molasses I can only say thank you, but not thank you. Other people feel different, so we can agree to disagree. E.g. why the hell the new jmoll graphics does not work on my Fedora 7 install from the sage command line? Some of this might be due to the quickness with which jmol was added to Sage so that it could be demonstrated at the AMS meeting. I suspect interaction between gcc's libgjc and jmol. libgjc is close to Sun's java in compability, but not equal. Installing and using Sun's JDK 1.6 might fix the issue. Ted Cheers, Michael I really don't want to start a flame war about this Abshoff --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~--~~~~--~~--~--~---
[sage-devel] Re: Wiris -- something like the Sage notebook sort of
On Jan 8, 2008, at 4:31 PM, Ted Kosan wrote: Jaap wrote: With bandwidth and memory capacity increasing exponentially, this is fast becoming a non-issue. I've heard that before! But I don't believe this is the panacea! The exponential growth of computing indicates that by 2020, a $1000 computer will have the same capacity as a human brain. By 2050, a $1000 computer will have the same capacity as all the human brains on the planet. You may be confusing capacity with capability :-} Regardless of the capacity of whatever shows up in the morning, these things will still be programmed by humans, who are notoriously clueless about how the brain works. That means that neither the hardware nor the software will be any more efficient than they currently are, and in addition, the software is being increasingly developed by programmers who are relying on that capacity to cover up for a real lack of talent and understanding. That last isn't intended as a slight/insult. I mean that more and more people are getting hauled into the programming profession without any real understanding of the job. Java is so simple a child can use it is something that (while hyperbole) product development managers seem to believe. This is one of the reasons (but certainly not the only) that we don't see an overall increase in system performance to match what we see in our hardware components. Your comment that computer capacity is not going to be a problem for Java seems indicative of that mind-set. I don't think Moore's Law figures into this: hardware and software will continue to progress to meet each other's demands, and when one slows, the other will too. Justin -- Justin C. Walker, Curmudgeon-At-Large Director Institute for the Enhancement of the Director's Income Weaseling out of things is what separates us from the animals. Well, except the weasel. - Homer J Simpson --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~--~~~~--~~--~--~---
[sage-devel] Re: Wiris -- something like the Sage notebook sort of
Ted Kosan wrote: Jaap wrote: With bandwidth and memory capacity increasing exponentially, this is fast becoming a non-issue. I've heard that before! But I don't believe this is the panacea! The exponential growth of computing indicates that by 2020, a $1000 computer will have the same capacity as a human brain. By 2050, a $1000 computer will have the same capacity as all the human brains on the planet. http://en.wikipedia.org/wiki/Image:PPTExponentialGrowthof_Computing.jpg http://www.kurzweilai.net/meme/frame.html?main=/articles/art0134.html? To me, exponential growth seems more like magic than anything else, even if it is just simple science :-) If these projections are even close to being accurate then computer capacity is not going to be a problem for Java. I truly believe that the real issue we will be dealing with in the not so distant future is the best way to upload Sage into human brains. Count me off! Maybe java is popular, but ask your self why? Managers want it, but they usually don't know why! Java is one of the most popular languages for manager-free open source projects on sourceforge too: http://www.cs.berkeley.edu/~flab/languages.html People and lemmings have something in common! See the primaries in the USA. The best thing java has brought us is Jython. I agree, though, that Java has a significant amount of hype associated with it and this is definitely a distraction. Ok, I once thought java was the way to go, but now I'm a convert. I hate java. Truth be told, I like programming in Python better than Java when this is feasible. For rich cross-platform browser-based applications, however, it is difficult to beat. I'm not convinced as long Microsoft exists!? E.g. why the hell the new jmoll graphics does not work on my Fedora 7 install from the sage command line? Some of this might be due to the quickness with which jmol was added to Sage so that it could be demonstrated at the AMS meeting. Don't think so. It just demonstrates that java is not the ubiquitous platform to rely on! Jaap --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~--~~~~--~~--~--~---
[sage-devel] Re: Wiris -- something like the Sage notebook sort of
Justin C. Walker wrote: On Jan 8, 2008, at 4:31 PM, Ted Kosan wrote: Jaap wrote: With bandwidth and memory capacity increasing exponentially, this is fast becoming a non-issue. I've heard that before! But I don't believe this is the panacea! The exponential growth of computing indicates that by 2020, a $1000 computer will have the same capacity as a human brain. By 2050, a $1000 computer will have the same capacity as all the human brains on the planet. You may be confusing capacity with capability :-} Regardless of the capacity of whatever shows up in the morning, these things will still be programmed by humans, who are notoriously clueless about how the brain works. That means that neither the hardware nor the software will be any more efficient than they currently are, and in addition, the software is being increasingly developed by programmers who are relying on that capacity to cover up for a real lack of talent and understanding. That last isn't intended as a slight/insult. I mean that more and more people are getting hauled into the programming profession without any real understanding of the job. Java is so simple a child can use it is something that (while hyperbole) product development managers seem to believe. Java simple? Teaching experience tells me that learning java is only possible under the condition that you know java! [..] Jaap --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~--~~~~--~~--~--~---
[sage-devel] Re: Wiris -- something like the Sage notebook sort of
This is a fun thread :-) Anyway, what I am more interested in than specific Sage client types is a good way for all client types to talk with Sage. Here is Wiris's communication architecture: http://www.wiris.com/images/stories/architecture_en.jpg What I would like to have in Sage is a protocol like what Wiris seems to have which is general enough to be used even with something out of the ordinary like educational portal software. I would like to use the protocol to allow Java applets to communicate with Sage, but I would also like to use it to allow OpenOffice to communicate with Sage directly and SecondLife to communicate with it using its scripting language. Any ideas on what a protocol like this would look like? Ted --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~--~~~~--~~--~--~---
[sage-devel] FLINT 1.05 make check failure on Linux/Itanium
Hello, Sage 2.10.alpha0 contains an updated FLINT release. As per usual I forced to run make check and the following happened in 64 bit mode with gcc 4.2.2 on Itanium: Testing _fmpz_poly_mul_KS()... GNU MP: Cannot reallocate memory (old_size=8 new_size=4294959128) ./spkg-check: line 51: 13014 Aborted ./fmpz_poly-test real13m45.435s user13m40.921s sys 0m1.597s sage: An error occurred while installing flint-1.05 Please email sage-devel http://groups.google.com/group/sage-devel explaining the problem and send the relevant part of of /users/wstein/build/sage-2.10.alpha0/install.log. Describe your computer, operating system, etc. If you want to try to fix the problem, yourself *don't* just cd to /users/wstein/build/sage-2.10.alpha0/spkg/build/flint-1.05 and type 'make'. Instead type /users/wstein/build/sage-2.10.alpha0/sage -sh in order to set all environment variables correctly, then cd to /users/wstein/build/sage-2.10.alpha0/spkg/build/flint-1.05 (When you are done debugging, you can type exit to leave the subshell.) make[1]: *** [installed/flint-1.05] Error 1 make[1]: Leaving directory `/scratch/was/build/sage-2.10.alpha0/spkg' real48m8.044s user45m32.578s sys 1m34.541s The gmp I am linking against passes make check. Obviously anybody else seeing failures on any platforms with the updated FLINT.spkg should report them here. Cheers, Michael --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~--~~~~--~~--~--~---
[sage-devel] Re: Wiris -- something like the Sage notebook sort of
On Jan 8, 2008 5:34 PM, Ted Kosan [EMAIL PROTECTED] wrote: This is a fun thread :-) Anyway, what I am more interested in than specific Sage client types is a good way for all client types to talk with Sage. Here is Wiris's communication architecture: http://www.wiris.com/images/stories/architecture_en.jpg What I would like to have in Sage is a protocol like what Wiris seems to have which is general enough to be used even with something out of the ordinary like educational portal software. I would like to use the protocol to allow Java applets to communicate with Sage, but I would also like to use it to allow OpenOffice to communicate with Sage directly and SecondLife to communicate with it using its scripting language. Any ideas on what a protocol like this would look like? Just to break up the tension in this thread a little bit, here's my idea of what it might look like: http://sage.math.washington.edu/home/was/tmp/architecture_en.jpg William --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~--~~~~--~~--~--~---
[sage-devel] Re: Wiris -- something like the Sage notebook sort of
On Tue, 8 Jan 2008, Ted Kosan wrote: Jaap wrote: Ok, I once thought java was the way to go, but now I'm a convert. I hate java. Truth be told, I like programming in Python better than Java when this is feasible. For rich cross-platform browser-based applications, however, it is difficult to beat. Java is for dinosaurs. Flash for the win! You can make pretty loading bars, *super* snazzy interfaces, and it's easy to use. Also, it can easily play sound (which the javascript notebook will probably not be capable of, ever). Flash has much better cross-platform support; it's a true web technology. --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~--~~~~--~~--~--~---
[sage-devel] Re: Wiris -- something like the Sage notebook sort of
On Jan 9, 5:39 am, [EMAIL PROTECTED] wrote: On Tue, 8 Jan 2008, Ted Kosan wrote: Jaap wrote: Ok, I once thought java was the way to go, but now I'm a convert. I hate java. Truth be told, I like programming in Python better than Java when this is feasible. For rich cross-platform browser-based applications, however, it is difficult to beat. Java is for dinosaurs. Flash for the win! You can make pretty loading bars, *super* snazzy interfaces, and it's easy to use. Also, it can easily play sound (which the javascript notebook will probably not be capable of, ever). Flash has much better cross-platform support; it's a true web technology. I beg to differ. I am not sure if my sarcasm detector is broken, but you seem to be serious. Ever installed flash on a 64 bit linux desktop or a PPC linux desktop? Flash sucks performance wise even more than Java. If I use YouTube or some other flash based video player it eats 20-30 percent of my CPU time with a postage stamp sized window without post processing or any other filter while XVID+MP3 hardware accelerated video via mplayer with 623x352 *and* post processing eats 0.3% CPU. We are talking about 2 orders of magnitude of CPU usage here without taking quality into consideration. With that it could be close to 3 orders of magnitude. Cheers, Michael --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~--~~~~--~~--~--~---
[sage-devel] Re: Wiris -- something like the Sage notebook sort of
On Jan 8, 2008 8:48 PM, mabshoff [EMAIL PROTECTED] wrote: On Jan 9, 5:39 am, [EMAIL PROTECTED] wrote: On Tue, 8 Jan 2008, Ted Kosan wrote: Jaap wrote: Ok, I once thought java was the way to go, but now I'm a convert. I hate java. Truth be told, I like programming in Python better than Java when this is feasible. For rich cross-platform browser-based applications, however, it is difficult to beat. Java is for dinosaurs. Flash for the win! You can make pretty loading bars, *super* snazzy interfaces, and it's easy to use. Also, it can easily play sound (which the javascript notebook will probably not be capable of, ever). Flash has much better cross-platform support; it's a true web technology. I beg to differ. I am not sure if my sarcasm detector is broken, but you seem to be serious. Ever installed flash on a 64 bit linux desktop or a PPC linux desktop? Flash sucks performance wise even more than Java. If I use YouTube or some other flash based video player it eats 20-30 percent of my CPU time with a postage stamp sized window without post processing or any other filter while XVID+MP3 hardware accelerated video via mplayer with 623x352 *and* post processing eats 0.3% CPU. We are talking about 2 orders of magnitude of CPU usage here without taking quality into consideration. With that it could be close to 3 orders of magnitude. Just a quick question: Isn't Flash a closed-source commercial product, hence completely unsuitable for use as a core technology in Sage? (In contrast, Java is (supposed to be?) GPL'd now. ) -- William --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~--~~~~--~~--~--~---
[sage-devel] Re: Wiris -- something like the Sage notebook sort of
On Jan 8, 2008 11:51 PM, William Stein [EMAIL PROTECTED] wrote: Just a quick question: Isn't Flash a closed-source commercial product, hence completely unsuitable for use as a core technology in Sage? (In contrast, Java is (supposed to be?) GPL'd now. ) -- William Yes it is closed (btw, I don't think Tom was advocating the use of Flash in Sage, just pointing out the fact that it is near-perfect for web developement in his view). I just want to point out the obvious about Javascript: it's a bona fide platform for web development that is installed by default on most modern browsers. You actually have to install Java (now that it's GPL-compatible) and you don't gain much in the short run for web dev. If you want to get serious about web development why not ship PHP, which was specifically written for the web? So it's -1 from me on Java *for web dev*. I would not discount Java for other things: a combination of Java platform + scripting language (like python, php, perl) could open up exciting possibilities for Sage . didier --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~--~~~~--~~--~--~---
[sage-devel] Re: Wiris -- something like the Sage notebook sort of
William wrote: Just to break up the tension in this thread a little bit, here's my idea of what it might look like: http://sage.math.washington.edu/home/was/tmp/architecture_en.jpg The idea of separating the Sage computation engine from the notebook server looks interesting. I took that idea, expanded upon it and then added standard web service protocols (http://en.wikipedia.org/wiki/Web_service): http://sage.math.washington.edu/home/tkosan/misc/web_service_architecture.png Ted --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~--~~~~--~~--~--~---
[sage-devel] Re: Wiris -- something like the Sage notebook sort of
On Tue, 8 Jan 2008, William Stein wrote: On Jan 8, 2008 8:48 PM, mabshoff [EMAIL PROTECTED] wrote: On Jan 9, 5:39 am, [EMAIL PROTECTED] wrote: On Tue, 8 Jan 2008, Ted Kosan wrote: Jaap wrote: Ok, I once thought java was the way to go, but now I'm a convert. I hate java. Truth be told, I like programming in Python better than Java when this is feasible. For rich cross-platform browser-based applications, however, it is difficult to beat. Java is for dinosaurs. Flash for the win! You can make pretty loading bars, *super* snazzy interfaces, and it's easy to use. Also, it can easily play sound (which the javascript notebook will probably not be capable of, ever). Flash has much better cross-platform support; it's a true web technology. I beg to differ. I am not sure if my sarcasm detector is broken, but you seem to be serious. Ever installed flash on a 64 bit linux desktop or a PPC linux desktop? Flash sucks performance wise even more than Java. If I use YouTube or some other flash based video player it eats 20-30 percent of my CPU time with a postage stamp sized window without post processing or any other filter while XVID+MP3 hardware accelerated video via mplayer with 623x352 *and* post processing eats 0.3% CPU. We are talking about 2 orders of magnitude of CPU usage here without taking quality into consideration. With that it could be close to 3 orders of magnitude. Just a quick question: Isn't Flash a closed-source commercial product, hence completely unsuitable for use as a core technology in Sage? (In contrast, Java is (supposed to be?) GPL'd now. ) Yes, and as Michael almost detected, I was being sarcastic. Didier is nearly right -- I think flash is neat, and in some ways better than Java: relevantly, for web-based software. Here's the truth, though. Memory and time concerns regarding Java are nearly insignificant when compared to javascript. However, javascript is nearly platform-independent (though there are a number of freakish differences, the gap is diminishing), and it is a *very* active area. The amount of Java on the web has remained nearly constant since the late 90's, support has gotten better in places, and worse in others. Everybody has their own opinion on what technology is best, and they all have great justifications. The notebook works, and it's being improved. If Ted wants to write a Java notebook, cool. If Timothy wants to write a Flash notebook, cool. I'd really prefer people to help out with the notebook, since homework already keeps me stretched thin, and I'd rather be writing more mathematical code. But hey, y'all are volunteers -- do what you want. I'll complain if the interface design sucks, and try to offer constructive advice. --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~--~~~~--~~--~--~---
[sage-devel] Re: Block matrices?
On Jan 8, 2008 4:27 AM, vgermrk [EMAIL PROTECTED] wrote: Is there a way to construct block matrices in SAGE? Not just the block_sum, augment and stack functions. As an example, let A, B, C, D be matrices and i want to construct a matrix like E=[[A,B],[C,D]] Such a feature would be very nice. Sage's MatrixSpace and matrix don't have support for this. numpy (which you get via import numpy) might have support for numerical matrices like this. --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~--~~~~--~~--~--~---
[sage-devel] Fwd: Sage Mac application bundle
Hi, Can some Mac user try this code submission from NASA out and make it part of Sage? It looks pretty sweet. -- Forwarded message -- From: Ted Wright @ NASA Date: Jan 8, 2008 10:44 AM Subject: Sage Mac application bundle To: [EMAIL PROTECTED] Hi, Thanks for Sage - it's awesome. I need to convince my coworkers to switch from their proprietary programs to Sage. I've attached a little script that uses the Platypus program (http://www.sveinbjorn.org/software) to bundle the sage directory into a clickable Mac application. It has some code to update the SAGE_ROOT variable so that things still work if a user drags the program around. My code is public domain, so feel free to use it if you like it. Ted -- William Stein Associate Professor of Mathematics University of Washington http://wstein.org --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~--~~~~--~~--~--~--- sageMac.tgz Description: GNU Zip compressed data
[sage-devel] Re: Sage Mac application bundle
OK, I've made this sage mac app thing trac #1731: http://trac.sagemath.org/sage_trac/ticket/1731 On Jan 8, 2008 10:45 PM, William Stein [EMAIL PROTECTED] wrote: Hi, Can some Mac user try this code submission from NASA out and make it part of Sage? It looks pretty sweet. -- Forwarded message -- From: Ted Wright @ NASA Date: Jan 8, 2008 10:44 AM Subject: Sage Mac application bundle To: [EMAIL PROTECTED] Hi, Thanks for Sage - it's awesome. I need to convince my coworkers to switch from their proprietary programs to Sage. I've attached a little script that uses the Platypus program (http://www.sveinbjorn.org/software) to bundle the sage directory into a clickable Mac application. It has some code to update the SAGE_ROOT variable so that things still work if a user drags the program around. My code is public domain, so feel free to use it if you like it. Ted -- William Stein Associate Professor of Mathematics University of Washington http://wstein.org -- William Stein Associate Professor of Mathematics University of Washington http://wstein.org --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~--~~~~--~~--~--~---
[sage-devel] Re: Block matrices?
On Jan 9, 7:43 am, William Stein [EMAIL PROTECTED] wrote: On Jan 8, 2008 4:27 AM, vgermrk [EMAIL PROTECTED] wrote: Is there a way to construct block matrices in SAGE? Not just the block_sum, augment and stack functions. As an example, let A, B, C, D be matrices and i want to construct a matrix like E=[[A,B],[C,D]] Such a feature would be very nice. Sage's MatrixSpace and matrix don't have support for this. numpy (which you get via import numpy) might have support for numerical matrices like this. CVXOPT also has support for this: http://abel.ee.ucla.edu/cvxopt/examples/short-examples/creating-matrices/ Joachim --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~--~~~~--~~--~--~---
[sage-devel] Re: Block matrices?
On Jan 8, 2008 11:05 PM, [EMAIL PROTECTED] [EMAIL PROTECTED] wrote:
On Jan 9, 7:43 am, William Stein [EMAIL PROTECTED] wrote:
On Jan 8, 2008 4:27 AM, vgermrk [EMAIL PROTECTED] wrote:
Is there a way to construct block matrices in SAGE?
Not just the block_sum, augment and stack functions.
As an example, let A, B, C, D be matrices and i want to construct a
matrix like E=[[A,B],[C,D]]
Such a feature would be very nice.
Sage's MatrixSpace and matrix don't have support for this.
numpy (which you get via import numpy) might have support
for numerical matrices like this.
CVXOPT also has support for this:
http://abel.ee.ucla.edu/cvxopt/examples/short-examples/creating-matrices/
And, just to be clear, CVXOPT is standard in Sage.
There are a number of caveats though. Currently you really have to switch
to python mode to create a cvxopt matrix in Sage, probably because of
preparser issues (see below). Also, more importantly, it looks to me
like making a matrix from matrices in CVXOPT does *not* make a matrix
with matrix entries, but instead makes a single matrix -- see below.
{{{id=0|
%python
from cvxopt.base import matrix
A = matrix([1.0, 2.0, 3.0, 4.0, 5.0, 6.0], (2,3))
print A
///
1.e+00 3.e+00 5.e+00
2.e+00 4.e+00 6.e+00
}}}
{{{id=1|
%python
B = matrix([ [1.0, 2.0], [3.0, 4.0] ])
print B
///
1.e+00 3.e+00
2.e+00 4.e+00
}}}
{{{id=2|
%python
a = matrix([ [A] ,[B] ])
}}}
{{{id=4|
print a
///
1.e+00 3.e+00 5.e+00 1.e+00 3.e+00
2.e+00 4.e+00 6.e+00 2.e+00 4.e+00
}}}
--~--~-~--~~~---~--~~
To post to this group, send email to sage-devel@googlegroups.com
To unsubscribe from this group, send email to [EMAIL PROTECTED]
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/
-~--~~~~--~~--~--~---
[sage-devel] Re: Anyone knows about any regression in gmp-4.2.2?
On Jan 9, 10:00 am, mabshoff [EMAIL PROTECTED] dortmund.de wrote: On Jan 8, 9:55 pm, Francois [EMAIL PROTECTED] wrote: On Jan 9, 8:58 am, mabshoff [EMAIL PROTECTED] dortmund.de wrote: Ok, we would certainly welcome any kind of participation from the Gentoo community. Hopefully some support may be drummed there, Excellent. I myself may be soon claimed by a non-academic real-life that could put me on hiatus. Sooner or later it happens to most of us ;) Very funny. I would have preferred the later option. Which kinds of bring to another thing that I wanted to ask about for a while. I noticed that there is in fact 2 blas packages in sage, atlas and what is usually refereed as blas-reference in Gentoo. It seems that the goal of that one is just to provide f77blas. Do we *really* need that? It depends. We need to build Lapack before ATLAS since we build an improved liblapack with the Lapack bits from ATLAS. Now if we were to run Lapack's test suite we need a BLAS. Atlas is also not standard yet on OSX, so there are more considerations here. Since netlib.org's BLAS is small and compiles quickly nobody has advocated its removal and since I consider it important to be able to run the Lapack testers on demand I would like it to stay. Ok make sense, it's one thing that I had disabled without much side effect apart from touching cvxopt about lf77blas although I believe numpy, scipy and al will detect and use it if present. Plus while building with my system atlas it failed because it wanted to remove /usr/lib/libblas.so. Something to think about for your SYSTEM_ATLAS setting. Ok, that does seem an odd thing to do. Josh told me a while back that numpy, scipy and cvxopt all use atlas per default, so maybe there is something different the way ATLAS is shipped on Gentoo. Sooner or later I will remove F77 BLAS and see what happens ;) Well there is some stuff in Gentoo in that we can install several version of blas, lapack and cblas and switch between them at will. However the python package you mention indeed detect atlas more or less automatically. Reviewing what I have done and why: *linbox, not really mentioned before but took a little tweak in the atlas department: set LINBOX_BLAS=yes in deps and it will take care of itself. *numpy: I set the variables ATLAS, BLAS and LAPACK in deps, it didn't seem to find them otherwise. I wasn't aware of the SAGE_ATLAS variable at the time - it may be enough to set it. *scipy/scipy_sandbox: most frustrating, the method employed for numpy didn't work because spkg-install explicitly define those variables unless you are on OS X in which case they are unset. *cvxopt: add to touch because unless your are on OS X spkg-install wants it to compile with -lf77blas which is of course provided by something I had removed. Looking at the Gentoo ebuild the BLAS, LAPACK and ATLAS variables are set to None and the output of a query to pkg-config is fed into site.cfg (that's for numpy - nothing special is done for scipy although it says there can be linking problems with lapack and checks have to be done for each version bumps). --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~--~~~~--~~--~--~---
[sage-devel] Re: Block matrices?
Sage's MatrixSpace and matrix don't have support for this. numpy (which you get via import numpy) might have support for numerical matrices like this. CVXOPT also has support for this: http://abel.ee.ucla.edu/cvxopt/examples/short-examples/creating-matri... And, just to be clear, CVXOPT is standard in Sage. There are a number of caveats though. Currently you really have to switch to python mode to create a cvxopt matrix in Sage, probably because of preparser issues (see below). Also, more importantly, it looks to me like making a matrix from matrices in CVXOPT does *not* make a matrix with matrix entries, but instead makes a single matrix -- see below. That is true; CVXOPT only supports 2-D numerical matrices, whereas Numpy supports N-D arrays of general objects. --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~--~~~~--~~--~--~---
[sage-devel] Re: Block matrices?
It was my impression that he didn't want a matrix with matrix entries,
but instead wanted the matrix whose entries were given by the entries
of the submatrices.
--Mike
On Jan 8, 2008 11:12 PM, William Stein [EMAIL PROTECTED] wrote:
On Jan 8, 2008 11:05 PM, [EMAIL PROTECTED] [EMAIL PROTECTED] wrote:
On Jan 9, 7:43 am, William Stein [EMAIL PROTECTED] wrote:
On Jan 8, 2008 4:27 AM, vgermrk [EMAIL PROTECTED] wrote:
Is there a way to construct block matrices in SAGE?
Not just the block_sum, augment and stack functions.
As an example, let A, B, C, D be matrices and i want to construct a
matrix like E=[[A,B],[C,D]]
Such a feature would be very nice.
Sage's MatrixSpace and matrix don't have support for this.
numpy (which you get via import numpy) might have support
for numerical matrices like this.
CVXOPT also has support for this:
http://abel.ee.ucla.edu/cvxopt/examples/short-examples/creating-matrices/
And, just to be clear, CVXOPT is standard in Sage.
There are a number of caveats though. Currently you really have to switch
to python mode to create a cvxopt matrix in Sage, probably because of
preparser issues (see below). Also, more importantly, it looks to me
like making a matrix from matrices in CVXOPT does *not* make a matrix
with matrix entries, but instead makes a single matrix -- see below.
{{{id=0|
%python
from cvxopt.base import matrix
A = matrix([1.0, 2.0, 3.0, 4.0, 5.0, 6.0], (2,3))
print A
///
1.e+00 3.e+00 5.e+00
2.e+00 4.e+00 6.e+00
}}}
{{{id=1|
%python
B = matrix([ [1.0, 2.0], [3.0, 4.0] ])
print B
///
1.e+00 3.e+00
2.e+00 4.e+00
}}}
{{{id=2|
%python
a = matrix([ [A] ,[B] ])
}}}
{{{id=4|
print a
///
1.e+00 3.e+00 5.e+00 1.e+00 3.e+00
2.e+00 4.e+00 6.e+00 2.e+00 4.e+00
}}}
--~--~-~--~~~---~--~~
To post to this group, send email to sage-devel@googlegroups.com
To unsubscribe from this group, send email to [EMAIL PROTECTED]
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/
-~--~~~~--~~--~--~---
[sage-devel] Re: Block matrices?
That will be really easy to implement, I'll do it right now.
- Robert
On Jan 8, 2008, at 11:25 PM, Mike Hansen wrote:
It was my impression that he didn't want a matrix with matrix entries,
but instead wanted the matrix whose entries were given by the entries
of the submatrices.
--Mike
On Jan 8, 2008 11:12 PM, William Stein [EMAIL PROTECTED] wrote:
On Jan 8, 2008 11:05 PM, [EMAIL PROTECTED]
[EMAIL PROTECTED] wrote:
On Jan 9, 7:43 am, William Stein [EMAIL PROTECTED] wrote:
On Jan 8, 2008 4:27 AM, vgermrk [EMAIL PROTECTED] wrote:
Is there a way to construct block matrices in SAGE?
Not just the block_sum, augment and stack functions.
As an example, let A, B, C, D be matrices and i want to
construct a
matrix like E=[[A,B],[C,D]]
Such a feature would be very nice.
Sage's MatrixSpace and matrix don't have support for this.
numpy (which you get via import numpy) might have support
for numerical matrices like this.
CVXOPT also has support for this:
http://abel.ee.ucla.edu/cvxopt/examples/short-examples/creating-
matrices/
And, just to be clear, CVXOPT is standard in Sage.
There are a number of caveats though. Currently you really have
to switch
to python mode to create a cvxopt matrix in Sage, probably because of
preparser issues (see below). Also, more importantly, it looks to me
like making a matrix from matrices in CVXOPT does *not* make a matrix
with matrix entries, but instead makes a single matrix -- see below.
{{{id=0|
%python
from cvxopt.base import matrix
A = matrix([1.0, 2.0, 3.0, 4.0, 5.0, 6.0], (2,3))
print A
///
1.e+00 3.e+00 5.e+00
2.e+00 4.e+00 6.e+00
}}}
{{{id=1|
%python
B = matrix([ [1.0, 2.0], [3.0, 4.0] ])
print B
///
1.e+00 3.e+00
2.e+00 4.e+00
}}}
{{{id=2|
%python
a = matrix([ [A] ,[B] ])
}}}
{{{id=4|
print a
///
1.e+00 3.e+00 5.e+00 1.e+00 3.e+00
2.e+00 4.e+00 6.e+00 2.e+00 4.e+00
}}}
--~--~-~--~~~---~--~~
To post to this group, send email to sage-devel@googlegroups.com
To unsubscribe from this group, send email to [EMAIL PROTECTED]
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/
-~--~~~~--~~--~--~---
[sage-devel] Re: Fwd: these are the links I mentioned...
Very cool. I'm interested, at some magical future point when I have time, in pushing Sage on supercomputers. I have access to various large systems at the Minnesota Supercomputing Institute, I just haven't had time to pursue it. I'll have to learn more about Vision, it looks like there are some useful lessons there. Is anyone else as bothered by Tk as I am? To me, it always looks outdated without any sort of cool retro feel, quite a trick. -Marshall On Jan 8, 3:31 pm, Fernando Perez [EMAIL PROTECTED] wrote: On Jan 8, 2008 4:18 PM, William Stein [EMAIL PROTECTED] wrote: Sage devel might be interested in this -- it's screencasts of a tool that numerical python people that DoD / supercomputing sorts have developed Wow, this is cool. We have an ipython1 screencast now :) I'd never heard of this... Vision is really neat, BTW, Developed at the Scripps institute by a very good team led by Michel Sanner. They've presented at various scipy conferences: http://www.scripps.edu/~sanner/ Cheers, f --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~--~~~~--~~--~--~---
