Update according to @lamblin suggestions.

doc/install_generic.inc

@@ -9,7 +9,23 @@ Installation
 Stable Installation
 -------------------
 
-.. comment: install_theano_stable_version
+With ``conda``
+^^^^^^^^^^^^^^
+
+If you use conda, you can directly install both theano and pygpu. Libgpuarray
+will be automatically installed as a dependency.
+
+.. code-block:: bash
+
+    conda install theano pygpu
+
+With ``pip``
+^^^^^^^^^^^^
+
+If you use pip, you have to install Theano and libgpuarray separately.
+
+theano
+::::::
 
 Install the latest stable version of Theano with:
 
@@ -31,12 +47,8 @@ If you encountered any trouble, head to the :ref:`troubleshooting` page.
 
 The latest stable version of Theano is ``0.9.0`` (tagged with ``rel-0.9.0``).
 
-.. comment: end_install_theano_stable_version
-
 libgpuarray
-^^^^^^^^^^^
-
-.. comment: install_libgpuarray_stable_version
+:::::::::::
 
 For the stable version of Theano you need a specific version of libgpuarray,
 that has been tagged ``v0.6.2``.

doc/install_windows.txt

@@ -14,37 +14,21 @@ Windows Installation Instructions
 .. |PlatformCompiler| replace:: ``python-dev``, ``g++`` >= 4.2
 .. |CompilerName| replace:: ``g++``
 
-.. comment: Requirements, optional requirements and installation of miniconda.
+.. List of requirements, optional requirements, and installation of miniconda.
 .. include:: requirements.inc
-    :end-before: .. comment: install_requirements_and_optional_packages
+    :end-before: .. install_requirements_and_optional_packages
 
 Install requirements and optional packages
 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
 
-.. note::
-
-    * Arguments between <...> are optional.
-    * ``m2w64-toolchain`` package provides a fully-compatible version of GCC and is then highly recommended.
-
-For Python 2*
-:::::::::::::
-
 .. code-block:: bash
 
     conda install numpy scipy mkl-service libpython <m2w64-toolchain> <nose> <nose-parameterized> <sphinx> <pydot-ng>
 
-For Python 3*
-:::::::::::::
-
-.. code-block:: bash
-
-    conda install numpy scipy mkl-service libpython <m2w64-toolchain> <nose> <nose-parameterized> <sphinx>
-
-To install ``pydot`` (optional):
-
-.. code-block:: bash
+.. note::
 
-    pip install pydot-ng
+    * Arguments between <...> are optional.
+    * ``m2w64-toolchain`` package provides a fully-compatible version of GCC and is then highly recommended.
 
 .. _gpu_windows:
 
@@ -63,41 +47,9 @@ to install the CUDA driver and the CUDA Toolkit.
 
 You must reboot the computer after the driver installation.
 
-Installation of Theano and libgpuarray
-======================================
-
-Stable Installation
--------------------
-
-With ``conda``
-^^^^^^^^^^^^^^
-
-If you use conda, you can directly install both theano and pygpu. Libgpuarray
-will be automatically installed as a dependency.
-
-.. code-block:: bash
-
-    conda install theano pygpu
-
-With ``pip``
-^^^^^^^^^^^^
-
-If you use pip, you have to install Theano and libgpuarray separately.
-
-theano
-::::::
-
-.. comment: Installation instructions for theano stable version.
-.. include:: install_generic.inc
-    :start-after: .. comment: install_theano_stable_version
-    :end-before: .. comment: end_install_theano_stable_version
-
-libgpuarray
-:::::::::::
-
-.. comment: Installation instructions (libgpuarray for stable version, then full bleeding-edge and developer version instructions).
+.. Installation of Theano and libgpuarray.
 .. include:: install_generic.inc
-    :start-after: .. comment: install_libgpuarray_stable_version
+    :start-after: .. _install_generic:
 
 Instructions for other Python distributions (not recommended)
 =============================================================

doc/internal/how_to_release.txt

@@ -153,7 +153,7 @@ For final releases, send the e-mail to the following mailing lists:
 * theano-users
 * theano-announce
 * numpy-discussion@scipy.org
-* scipy-user@scipy.org
+* scipy-user@python.org
 * G+, Scientific Python: https://plus.google.com/communities/108773711053400791849
 
 For release candidates, only e-mail:

doc/requirements.inc

@@ -14,7 +14,7 @@ Requirements
 .. _pycuda: https://mathema.tician.de/software/pycuda/
 .. _skcuda: http://scikit-cuda.readthedocs.io/en/latest/
 
-    Python_ == 2.7 or >= 3.3
+    Python_ == 2.7 or ( >= 3.3 and <= 3.5 )
         The development package (python-dev or
         python-devel on most Linux distributions) is recommended (see
         just below). Python 2.4 was supported up to and including the
@@ -71,7 +71,7 @@ Follow this `link <http://conda.pydata.org/miniconda.html>`__ to install Minicon
 
     If you want fast compiled code (recommended), make sure you have |CompilerName| installed.
 
-.. comment: install_requirements_and_optional_packages
+.. install_requirements_and_optional_packages
 
 Install requirements and optional packages
 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^