On Tue, Jun 3, 2014 at 1:53 PM, Olof Bjarnason <olof.bjarna...@gmail.com> wrote:
> Why isn't CPython+NumPy+SciPy (or what you need on top of NumPy) > enough? It's been tested and maintained for a long time, and works > quite well? > For standalone stuff, yes. But if I'm building an application in .NET that I want users to be able to script using IronPython, and those users might want to do some heavy-duty math (say, a 3D modeling application), it makes sense for numpy to be available directly to IronPython. > It does seem like a daunting task to try and build and maintain > something separate from the mainline NumPy/SciPy community... > Exactly, which is why I would prefer to see use work as closely as possible with them and the PyPy team (remembering that those two didn't work so closely in the early days of NumPyPy...) to reduce the workload. - Jeff
_______________________________________________ Ironpython-users mailing list Ironpython-users@python.org https://mail.python.org/mailman/listinfo/ironpython-users
