[Numpy-discussion] NEP-18 comment
Stefan van der Walt
stefanv at berkeley.edu
Wed Mar 6 17:22:14 EST 2019
It's great to have your input here, Frederic—thank you!
On Wed, 06 Mar 2019 12:41:11 -0800, Stephan Hoyer wrote:
> I agree, this seems like an easy and worthwhile change. The NEP-18
> implementation has been rewritten in C, so I expect that the typical
> overhead will be quite minimal (about 1 us per function call), but I agree
> that there may be some important edge cases that we missed.
The environment variable made sense when we only wanted to give select
access to try the new feature, but I think it's a slippery slope to
allow it to *disable* functionality that, for everyone else, is default.
Once you start having functionality vary across environments, you are
bound to see strange things happen. E.g., the same code, run with the
same verison of dask, may or may not work depending on a shell setting?
More information about the NumPy-Discussion