Set PYSPARK_PYTHON path in python2 kernelspec

(c) Copyright IBM Corp. 2016

(c) Copyright IBM Corp. 2016

* Install python2 spec to CONDA_DIR instead of system prefix too

(c) Copyright IBM Corp. 2016

(c) Copyright IBM Corp. 2016

Move to "internal" folder

(c) Copyright IBM Corp. 2016

Added ubuntu:14.04 locale fix

Bump kernel gateway to 0.3.1

Also change CMD vs ENTRYPOINT to match what minimal-notebook
and some major official Docker images do (redis, mysql, nginx, ...)
by having the jupyter kernelgateway in the CMD, and just tini --
in the entrypoint

(c) Copyright IBM Corp. 2016

Notebook 4.1.x

(c) Copyright IBM Corp. 2016

(c) Copyright IBM Corp. 2016

(c) Copyright IBM Corp. 2016

Also fix copyright headers on various files

Remove build deps from minimal-kernel image

(c) Copyright IBM Corp. 2016

* Use setuptools to get modern pip (debian packaged pip stops
  working after requests gets updated)
* Don't use wget to fetch tini, use Python!

(c) Copyright IBM Corp. 2016

* Remove SHELL env var and leftover USER nobody
* Install wget for tini then immediately uninstall it
* Install build-essentials and python3-dev for tornado speed up
  then immediately uninstall them

(c) Copyright IBM Corp. 2016

First time we've added a stack under the new versioning scheme

(c) Copyright IBM Corp. 2016

(c) Copyright IBM Corp. 2016

New docker stack for a headless kernel

Contribution (c) Copyright IBM Corp. 2016

Contribution (c) Copyright IBM Corp. 2016

Contribution (c) Copyright IBM Corp. 2016

Contribution (c) Copyright IBM Corp. 2016

Contribution (c) Copyright IBM Corp. 2016

Contribution (c) Copyright IBM Corp. 2015

New docker stack for a headless kernel.  It uses the new kernel gateway to provide websocket access without exposing the whole notebook UI.

Contribution (c) Copyright IBM Corp. 2015

Add texlive-generic-recommended to fix PDF export

Fixes #99

(c) Copyright IBM Corp. 2016

Make jovyan the default user for the docker cmd

Install libxrender1 to appease matplotlib et al.

Bite the bullet and preinstall it so that plotting libs that default
to using desktop rendering just work (matplotlib, ggplot, ...)
out of the box without having to get configuration right beforehand
(e.g., %matplotlib inline ahead of matplotlib import)

Only adds ~100k to the image size

(c) Copyright IBM Corp. 2015

make-deploy example: run stacks on docker-machines