Michael C. Grant

Oh! You know what, this is actually an issue with `nb_conda`, not `nb_conda_kernels`. I apologize for not realizing this before. I'll transfer this issue over.

I must be honest, I don't know if it's possible or not. I do know that the `conda info` information is cached every `CACHE_TIMEOUT` seconds but I can't recall the...

I am finding that `nb_conda_kernels` can indeed discover kernels if 1) you wait longer than 60 seconds from the environment creation and 2) you refresh the browser. I'm working a...

With JupyterLab, you don't have to refresh the browser—it will pick up new kernels in a minute or two, because JLab will periodically (about once every two minutes) query the...

I'm well aware that newer versions of the `jupyter_client` module are going to provide a more elegant method for allowing kernel providers to supply kernelspecs, rendering this approach obsolete. But...

We do this primarily for historical reasons. I'm happy to consider a PR to modify this behavior, but my preference would be to make it configurable so that it can...

Great!

> Today, nb_conda_kernels simply looks for the discovery module via import and, if not present (in jupyter_client), derives the CondaKernelProvider from object in order to satisfy loads knowing the find_kernels()...

I'm not concerned at all about 3. We will be able to support Python 2.7 with less functionality. Remember, if `nb_conda_kernels` is in Python 2.7 environment, it won't be able...

Ah, yes, you're right, they would indeed cause errors. We'll have to have separate async and non-async submodules.