On Aug 8, 2:56 am, "Alec Mihailovs" <[EMAIL PROTECTED]> wrote:
> From: "Bill Page" <[EMAIL PROTECTED]>
>
> > As technically hard as it might be, I think having a native Windows
> > version of Sage - even if it includes only a subset of the standard
> > packages - would likely be a big factor in attracting more users.
>

Having worked on the Cygwin Port up to the point where it was dropped
I can only tell you that it is a pain to get to run. At that
particular point there were problem with at least 6 packages, two of
which were unresolved with no real hope of getting resolved any time
soon.

Getting other packages to work with MSVC is even a bigger problem, I
do not think that many packages of SAGE besides NTL compile with MSVC.
MinGW is worst than Cygwin because it doesn't provide any extra POSIX
functionality. For example: Singular does require mmap for its
allocator omalloc (I believe it can be made to use malloc, but from my
last conversation with Michael Brickenstein this is not a good idea to
switch away from omalloc).

> Being a Windows user, I can't agree less. Also, the notebook running in IE 7
> would be much more attractive for many Windows users (including me) than in
> Firefox.
>
> > Having even a subset of Sage available as a
> > native Windows application would introduce many more users to Sage and
> > probably motivate some of them to install Linux in order to access the
> > full version.
>
> I always have few Linuxes installed, just for running programs (such as
> SAGE) that are not available in Windows. Still, it's not the same.
>
> > I think the best tool for building a native Windows version of Sage is
> > probably MSYS/MinGW which is really a cross-compiler and gnu tool set
> > that provides a Linux-like environment only during the build. The end
> > product is a native Windows application that does not depend on any
> > Linux emulation layer.

Cygwin is not Linux emulation, but provides a POSIX compatible layer
for Windows. It is not required for a user to install Cygwin, you just
need the right DLLs.

> >Unfortunately some of the standard packages in
> > Sage can not be built in this way and to make matters worse, as far as
> > I know the pexpect module that is required for interface with packages
> > like Maxima has not been successfully ported to Windows.
>
> However, for Python extensions, the compiler should be the same as the
> compiler used to build Python - for Windows it is Visual Studio (Express is
> OK) 2005.
>

There is some other code around to create pseudo TTYs for python on
Windows which are required for pexpect.

> Alec


--~--~---------~--~----~------------~-------~--~----~
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/
-~----------~----~----~----~------~----~------~--~---

Reply via email to