[Numpy-discussion] Best way to run python parallel
Thu Mar 29 10:44:22 CDT 2007
On 3/29/07, Steven H. Rogers <email@example.com> wrote:
> Another option coming 'Real Soon Now` is `IPython1
> <http://ipython.scipy.org/moin/IPython1>`. It's the enhanced Python
> interactive shell with the added capability for distributed interactive
This is moving slower than we'd all like due to the many constraints
on our time, but it is being actively developed.
I'm planning on releasing ipython 0.7.4 in the next week or two, and
Brian is hard at work cleaning up the 'saw' branch where all recent
development has taken place.
Once 0.7.4 (of trunk) is out, we'll bring all that code into saw to
begin full integration.
On April 28, we're having a coding sprint in Boulder, Colorado, to
work on this with other interested parties from the Colorado Front
I hope that sprint will both serve as a motivating deadline for me,
and as a good opportunity to move forward and recruit new
In summary, I hope that by May things may have stabilized at least to
the point that:
1. The saw branch is the only one actively developed, and will be
usable for day to day production work, albeit with the caveats of a
2. We'll have some simple examples/demos on how to use it included in
3. That branch will contain a 'normal' ipython as well, though I don't
expect that one to replace trunk yet.
More information about the Numpy-discussion