Mathics on Windows

[Jan Pöschko]

Hi all,

Did anybody ever try installing Mathics on Windows?

Thanks,

Jan

Begin forwarded message:

From: Gilberto Garcia-Pulgarin <garciap...@gmail.com>

Date: September 21, 2012 8:48:47 PM GMT+02:00

To: j...@poeschko.com

Subject: Hi, from Colombia

Reply-To: giga...@ciencias.udea.edu.co

Hello Jan,

I want to use mathics online or download it to mi PC, with windows, but I was unlucky in both.

Can you send me some information.

Thanks and best regards from Colombia

[Enrique Castaneda]

I'll try and feed back. I was also thinking in a setup

[Enrique Castaneda]

So, I was reading about the windows installation of GMP and found this (http://stackoverflow.com/questions/4711315/build-gmp-on-64bit-windows):

See http://gladman.plushost.co.uk/oldsite/computing/gmp4win.php , and note this paragraph;

GMP cannot, however, provide high performance 64-bit libraries on Windows because its 64-bit assembler code is incompatible with Windows x64 calling conventions. This means that 64-bit GMP libraries on WIndows are very slow when compared with MPIR, which has excellent 64-bit assembler support.

So you either disable assembly with ./configure --disable-assembly or use MPIR.

I also found a bit up in that same page:

You might want to take a look at MPIR. It is a fork of GMP that natively supports Visual Studio.

A MPIR installation will have the necessary libraries for Mathics to run fine? 

[David Meeker]

Mathics seems like a pretty neat tool.  I got it running on Windows, and I decided to toss everything into a self-installing executable.  I tried the installation on a few different Windows 7 machines and XP virtual machines, and it seems to run OK.  I've put the Windows self-installed executable I made up on my website at:

http://www.femm.info/wiki/Mathics/

--Dave

[Angus]

Excellent. We should think about putting something like this on mathics.org. Moreover, it would be really awesome if we could set something like this up to happen automatically (with $> python setup.py bdist for example).

[David Meeker]

I get the same issue with the Sierpinski triangle example as well.  The error messages are pretty cryptic, so it's hard for me to tell exactly why Mathics 0.6 doesn't like it.

An interesting data point is that the server at mathics.net returns the $Version string:
Mathics 0.6.0rc1 on PyPy 2.7.3 (2.1.0+dfsg-3, Sep 12 2013, 13:13:48) using Django 1.5.5, SymPy 0.7.3, mpmath 0.17
and does correctly render the Sierpinski triangle example, whereas the Windows version I rolled up returns:
Mathics 0.6.0rc1 on CPython 2.7.3 (default, Apr 10 2012, 23:31:26) [MSC v.1500 32 bit (Intel)] using Django 1.6.1, SymPy 0.7.3, mpmath 0.17
A significant difference is that I'm using CPython, whereas the mathics.net server is using PyPy.

So, I rolled up a new installer based on PyPy 2.2.1 rather than CPython. On my machine, the PyPy version runs all of the Gallery examples without throwing any errors.  I've posted the PyPy version at http://www.femm.info/wiki/Mathics06

--Dave

N.B.  I screened both the CPython and PyPy versions with AVG 2014 and Microsoft Security Essentials.  No red flags appeared.

On Monday, December 30, 2013 11:04:35 AM UTC-5, Ádám Hegedűs wrote:

The Sierpinkski triangle example is bugged for me too.

[Angus]

Working on a fix for the Serpinski bug. At first glance it looks like it might be a Sympy bug.