>
> that was some years ago -- I wonder how much use it's seen?

You can go a surprisingly long way with Python's built-in and stdlib
containers, so I'm not surprised it's not very widely used.

Exactly — what are the odds that list or deque performance is your bottleneck?

However, the barrier to entry for a third party package is still quite a bit higher than the Stdlib. So if a third party package that provides nothing but a performance boost isn’t used much — that doesn’t mean it wouldn’t be well-used if it were in the stdlib.

Note: I am not advocating anything— I haven’t looked at blist at all.


I don't know what wheels are supposed to change here.  You could already
build binary packages for Windows before wheels existed.  

Yes, but with a different ecosystem an no virtual environment support. Being able to pip install binary wheels does make things easier for Windows users. 


The problem as
I understand it is that you need a Windows machine (or VM) together with
the required set of compilers, and have to take the time to run the
builds.

Yup — still the case. Also with Mac and OS-X. Distributing a package with a compiled component is still a lot more work.

With conda-forge though, one could simply submit a recipe and have all
builds done automatically in the cloud.  Your users then have to use
conda (rather than pip and 

I’m a big fan of conda and conda-forge.

But a similar auto-build system could support binary wheels and pypi as well. And indeed, the scipy folks are doing just that. 

My point was that the infrastructure for delivering complied packaged is much better than it was even a few years ago.

-CHB


--

Christopher Barker, Ph.D.
Oceanographer

Emergency Response Division
NOAA/NOS/OR&R            (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker@noaa.gov