Raw kernel's are not started with the environment of the current conda environment.

[rchiodo]

Multiple users reporting a DLL load failure on either loading the kernel, or loading other packages.

I get this when loading numpy:

# %% Basic Imports...
---------------------------------------------------------------------------
ImportError                               Traceback (most recent call last)
~\AppData\Local\Continuum\miniconda3\envs\jupyter\lib\site-packages\numpy\core\__init__.py in 
     23 try:
---> 24     from . import multiarray
     25 except ImportError as exc:

~\AppData\Local\Continuum\miniconda3\envs\jupyter\lib\site-packages\numpy\core\multiarray.py in 
     13 
---> 14 from . import overrides
     15 from . import _multiarray_umath

~\AppData\Local\Continuum\miniconda3\envs\jupyter\lib\site-packages\numpy\core\overrides.py in 
      6 
----> 7 from numpy.core._multiarray_umath import (
      8     add_docstring, implement_array_function, _get_implementing_args)

ImportError: DLL load failed: The specified module could not be found.

[rchiodo]

Root cause seems to be that the environment from conda is not used when launching the kernel.

[IanMatthewHuff]

I think this is already done. Pulling to validate.

[IanMatthewHuff]

Validated.