Add docpage for external packages

We may want a page like this to increase exposure to other packages that use or
build on Dask.  The goal here is that when someone says "How do I use Dask to
solve X kind of problem" that we can send them to a subsection of this page for
a set of links to other downstream projects.

Creating such a comprehensive list is hard.  This could use help from others.

docs/source/external-packages.rst

@@ -0,0 +1,62 @@
+External Packages
+=================
+
+Dask work occurs in several different projects, each for different domains.
+This page lists some of those projects.
+
+Motivation
+----------
+
+Dask is used to parallelize a variety of other projects and domains.  By
+creating a culture of many subprojects we enable smaller communities to more
+easily build, package, and maintain functionality useful for their domain
+with minimal coordination with the core project.
+
+However, the lack of a central all-inclusive package also means that peripheral
+domain-specific projects don't get as much exposure.  This page attempts to fix
+this by providing some exposure in documentation.
+
+Machine Learning
+----------------
+
+-  Dask GLM: `docs <http://dask-glm.readthedocs.io/en/latest/>`_ and `source <https://github.com/dask/dask-glm>`_.  Generalized linear models like Logistic Regression on large datasets.
+-  Dask-SearchCV: `docs <http://dask-searchcv.readthedocs.io/en/latest/>`_ and `source <https://github.com/dask/dask-searchcv>`_.  Hyperparameter optimization of Scikit-Learn models
+-  Dask PatternSearch: `source <https://github.com/eriknw/dask-patternsearch>`_.  Gradient free search optimization
+
+
+Image Analysis
+--------------
+
+The `dask-image github organization <https://dask-image.github.io/>`_ collects
+many common subprojects together based off of dask.array.
+
+
+File Formats
+------------
+
+-  Fastparquet: `docs <http://fastparquet.readthedocs.io/en/latest/>`_ and `source <https://github.com/dask/fastparquet>`_.  Parquet format with NumPy, and Numba
+-  Parquet-Cpp with Arrow: `docs <https://arrow.apache.org/docs/python/parquet.html>`_ and `parquet source <https://github.com/apache/parquet-cpp>`_ and `arrow source <https://github.com/apache/arrow>`_
+
+
+Data Sources
+------------
+
+-  S3: `docs <http://s3fs.readthedocs.io/en/latest/>`_ and `source <https://github.com/dask/s3fs>`_.  Amazon's S3
+-  GCS: `docs <http://gcsfs.readthedocs.io/en/latest/>`_ and `source <https://github.com/martindurant/gcsfs>`_.  Google's cloud storage
+-  HDFS: `docs <http://hdfs3.readthedocs.io/en/latest/>`_ and `source <https://github.com/dask/hdfs3>`_.  Hadoop File System wrappers for Python
+
+
+Cluster Managers
+----------------
+
+-  Kubernetes: `source <https://github.com/martindurant/dask-kubernetes>`_
+-  YARN: `source <https://github.com/dask/dask-yarn>`_
+-  Marathon: `source <https://github.com/mrocklin/dask-marathon>`_
+-  DRMAA: `source <https://github.com/dask/dask-drmaa>`_
+-  EC2: `source <https://github.com/dask/dask-ec2>`_
+
+
+Other
+-----
+
+-  XArray: `docs <http://xarray.pydata.org/en/stable/>`_ and `source <https://github.com/pydata/xarray>`_ for labeled and indexed arrays

docs/source/index.rst

@@ -234,6 +234,7 @@ often a better choice.  If you are a *core developer*, then you should start her
 
 * :doc:`debugging`
 * :doc:`support`
+* :doc:`external-packages`
 * :doc:`changelog`
 * :doc:`presentations`
 * :doc:`develop`
@@ -251,6 +252,7 @@ often a better choice.  If you are a *core developer*, then you should start her
 
    debugging.rst
    support.rst
+   external-packages.rst
    changelog.rst
    presentations.rst
    develop.rst