explicitly install SQLAlchemy for both python2 and python3. also add libmysqlclient and python wrappers for the most popular uses

Document how to set a hashed password

(c) Copyright IBM Corp. 2016

For #210, fix a few other minor nits along the way

Already supported via the command line.
Update all the READMEs.

(c) Copyright IBM Corp. 2016

Update R version to 3.3* in r-notebook and datascience-notebook

Temporary workaround for #210

Pin libjpeg to v8

(c) Copyright IBM Corp. 2016

Update R version to 3.3*

Default matplotlib to inline backend

Reduces image size by ~120 MB

(c) Copyright IBM Corp. 2016

Fixes #235

(c) Copyright IBM Corp. 2016

Update pandas version to 0.18*

inherit pyspark from scipy-notebook

removes duplicate installation of python 2 env and Python packages.

There are very few packages in scipy not in pyspark, so removing the duplicate seems valuable.

Only move julia kernelspec

Make use of available environment variables

Support on-demand creation of mounted work directory

rather than all kernelspecs that might be in the user kernelspec dir

Fixes broken [[ ]] syntax when default sh is used

update sympy to 1.0

base-notebook defines environment variables for the Conda install path and the
notebook user. However, in some instances, these locations were still hardcoded.
Let’s use the variables instead.

Use `jq` from conda-forge

Various tuneups

Add liveliness check

Fixes #123

(c) Copyright IBM Corp. 2016

Ref #188

(c) Copyright IBM Corp. 2016

Ref #188

(c) Copyright IBM Corp. 2016

Prevent docker from creating the
~/.local/share/jupyter directory as rw for root
only. Otherwise notebook can’t create runtime/
and other import subduers

(c) Copyright IBM Corp. 2016

Some makefile cleanup too

(c) Copyright IBM Corp. 2016