Update to Omnia Linux anvil

.travis.yml

@@ -23,7 +23,7 @@ install:
   # The Dockerfile that defines the image that for the build environment is
   # available in this repo at devtools/omnia-build-box/Dockerfile
   - if [[ "$TRAVIS_OS_NAME" == "linux" ]]; then
-      docker pull jchodera/omnia-build-box:cuda${CUDA_SHORT_VERSION}-amd30-clang38;
+      docker pull jchodera/omnia-linux-anvil:texlive-amd30-cuda${CUDA_SHORT_VERSION};
     fi
 
 script:
@@ -37,7 +37,7 @@ script:
   - if [[ "$TRAVIS_OS_NAME" == "linux" ]]; then
 
         docker run -e UPLOAD -e BINSTAR_TOKEN -e TRAVIS_PULL_REQUEST
-            -t -i --rm -v `pwd`:/io jchodera/omnia-build-box:cuda${CUDA_SHORT_VERSION}-amd30-clang38
+            -t -i --rm -v `pwd`:/io jchodera/omnia-linux-anvil:texlive-amd30-cuda${CUDA_SHORT_VERSION}
             bash /io/devtools/docker-build.sh;
 
     elif [[ "$TRAVIS_OS_NAME" == "osx" ]]; then

README.md

@@ -2,22 +2,92 @@
 * Appveyor-CI `win` builds [![AppVeyor Build status](https://ci.appveyor.com/api/projects/status/fyjgl66t943tf2yg/branch/master?svg=true)](https://ci.appveyor.com/project/jchodera/conda-recipes/branch/master)
 
 
-omnia-md/conda-recipes
-----------------------
+# omnia-md/conda-recipes
 
 The recipes here create conda packages for scientific and numerical software components associated with the [`omnia`](http://omnia.md) project.
-The packages built from these recipes are shared with the community on [anaconda.org](https://anaconda.org/omnia).
+The packages built from these recipes are shared with the community on [anaconda.org](https://anaconda.org/omnia). 
+These packages also depend on the `conda-forge` conda channel.
 
-### Installing packages from omnia
+## Installing packages from omnia
 
-To install a package
+To install a package (`mdtraj` for example)
 ```bash
+# Set the channel priority behavior
+conda config --set channel_priority false
+# Add conda-forge as the lower priority (FILO format for priority)
+conda config --add channels conda-forge
 # Add the omnia channel
-$ conda config --add channels omnia
+conda config --add channels omnia
 # Install the 'mdtraj' a package
-$ conda install mdtraj
+conda install mdtraj
 ```
 
+The `channel_priority` behavior we configure gives priority to the highest version of the package over all channels.
+
+The default behavior (`channel_priority true`) pulls packages from the highest priority chanel, independent of package 
+version numbers. i.e. an older version package on a higher priority channel will be installed over a newer version 
+of the same package on a lower priority channel.
+
+### Important!
+
+These packages must use the old style of conda channel priority until omnia is entierly on conda-forge!
+The instructions below enforce the old channel priority format, please visit 
+[the conda documentations on channel priority](https://conda.io/docs/channels.html) 
+for more information.
+
+## Migration to conda-forge
+The Omnia project has started migrating to [`conda-forge`](https://conda-forge.github.io/). New packages that 
+ do not depend on OpenMM should be developed on `conda-forge` and exiting packages which do not depend on OpenMM 
+ should start migrating if possible.
+
+### Planed Migration Stages
+
+1. (**Current**) Update build image to CentOS 6 from CentOS 5
+
+    The base Docker image for linux builds will be updated to CentOS 6 with its new glibc. The base image is the 
+    `conda-forge` anvil, with some [custom addons](https://hub.docker.com/r/jchodera/omnia-linux-anvil/~/dockerfile/) 
+    to include things like the AMD SDK, TexLive, and CUDA for GPU builds. The updated version will ensure packages 
+    can work on the `conda-forge` platform which is CentOS 6 based.
+
+1. Stop building packages that appear in `conda-forge`
+
+    We want to minimize the amount of work we have to do as maintainers. To that end, we will stop building things 
+    which freely appear on `conda-forge` and maintained by someone other than us!
+    For reproducibility purposes, we will keep our previously compiled versions, but they will not longer be updated.  
+
+1. Begin migration of "nearly-pure Python" and non-OpenMM dependent packages
+
+   * Packages which do not depend on OpenMM and can be run on CPUs only should start migrating over to `conda-forge` 
+   in preparation for the total migration. 
+   * Packages which can compile with just the `conda-forge` linux-anvil should also start migrating.
+   * *We highly encourage devs of individual packages to start migrating now.* 
+   * **Once a package is on conda-forge, it should no longer depend packages from omnia!**
+
+1. Begin migration of `omnia` dependent (but not OpenMM dependent) packages to `conda-forge`
+
+   * As more of the `omnia` dependencies appear on `conda-forge`, more packages can start moving over.
+   * This stage can 
+
+1. Determine the appropriate way to build packages which require more than the `conda-forge` linux-anvil can provide
+
+   * The `conda-forge` linux-anvil does not support some things such as some LaTeX packages, AMD SDK, and CUDA files. 
+   * We will need to reach out to the conda-forge people to see what the best course of action is
+
+1. Move OpenMM to `conda-forge`
+
+    * This will highly depend on the previous point and will probably have the longest time frame.
+
+1. Move packages which depend on OpenMM to `conda-forge`
+
+1. Finish the move of all packages to `conda-forge`
+   * Also ensure that all former omnia packages can be installed without the `omnia` conda channel  
+
+1. Change this repo into an archive for reproducibility.
+
+### How to migrate to `conda-forge` (for existing packages)
+
+PLACEHOLDER
+
 ### Supported versions
 
 Python packages are built against latest two releases of python (3.5 and 3.6) and python 2.7.
@@ -31,16 +101,19 @@ Packages which have a binary dependency on [numpy](http://www.numpy.org/) are bu
 
 The recipes here are automatically built using [Travis-CI](https://travis-ci.org/) for `linux` and `osx` and [Appveyor-CI](http://www.appveyor.com/) for `win`.
 
-For `linux` builds, we use a modified version of the [Holy Build Box](http://phusion.github.io/holy-build-box/), available [here](https://github.com/omnia-md/omnia-build-box), to ensure that the packages are fully compatible across multiple linux distributions and versions.
+For `linux` builds, we use a modified version of the 
+[conda-forge linux-anvil](https://github.com/omnia-md/omnia-linux-anvil/blob/master/Dockerfile), 
+to ensure that the packages are fully compatible across multiple linux distributions and versions.
 This build image contains the additional tools:
-* [clang](http://clang.llvm.org/) 3.8.1
-* [TeXLive](https://www.tug.org/texlive/) 2015
+* [TeXLive](https://www.tug.org/texlive/) 2016
 * The [CUDA](https://developer.nvidia.com/cuda-toolkit) Toolkit version 8.0
 * The [AMD APP SDK](http://developer.amd.com/tools-and-sdks/opencl-zone/amd-accelerated-parallel-processing-app-sdk/) 3.0
 
+There is an additional image which has [clang](http://clang.llvm.org/) 3.8.1 
+
 To build a package yourself, run `conda build <package_name>`, or `./conda-build-all ./*` to build multiple packages across each of the supported python/numpy configurations.
 
-### Contributing a recipe
+### Contributing a recipe (this has not been updated to reflect the conda-forge changes)
 
 1. Fork this repo
 2. Add your `conda` recipe for building your package `packagename` in a subdirectory called `packagename`. Feel free to use other recipes here as examples.

appveyor.yml

@@ -57,10 +57,11 @@ environment:
 install:
   - "SET PATH=%PYTHON%;%PYTHON%\\Scripts;%PATH%"
   - set PATH=C:\MinGW\msys\1.0\bin;%PATH%
+  - conda update -yq --all
+  - conda config --set channel_priority false
   - conda config --add channels omnia
   - conda config --add channels salilab
-  - conda update -yq --all
-  - conda install -yq conda-build=2.1.5 jinja2 anaconda-client
+  - conda install -yq conda-build jinja2 anaconda-client
   - powershell .\\devtools\\appveyor\\missing-headers.ps1
   # conda-build for some inane reason, effectively puts C:\cygwin\bin at the
   # front of the PATH, ahead of everything else, regardless. See

conda-build-all

@@ -96,6 +96,7 @@ def main():
         help="Set the Python version used by conda build",
         metavar="PYTHON_VER",
         default='27,34,35,36',
+
     )
     p.add_argument(
         '--numpy',

coverage/meta.yaml

@@ -15,6 +15,7 @@ build:
   number: 0
   entry_points:
     - coverage = coverage.cmdline:main
+  skip: True
 
 requirements:
   build:

cvxpy/meta.yaml

@@ -4,6 +4,8 @@ package:
 
 build:
   number: 0
+  # No mingwpy on python 3.5+ which is needed by cvxp[t
+  skip: True # [win and (py35 or py36)]
 
 source:
   git_url: https://github.com/cvxgrp/cvxpy

devtools/docker-build.sh

@@ -1,19 +1,9 @@
 #!/bin/bash
 set -e
-# Activate Holy Build Box environment.
-source /hbb_exe/activate
-
-# Disable PYTHONPATH
-unset PYTHONPATH
-
 set -x
-curl -s -O https://repo.continuum.io/miniconda/Miniconda3-latest-Linux-x86_64.sh
-bash Miniconda3-latest-Linux-x86_64.sh -b -p /anaconda
-PATH=/opt/rh/devtoolset-2/root/usr/bin:/anaconda/bin:$PATH
+conda config --set channel_priority false
 conda config --add channels omnia
-conda install -yq conda-build=2.1.5 jinja2 anaconda-client
-# Install missing LaTeX docs
-tlmgr install fncychap tabulary capt-of eqparbox environ trimspaces
+conda install -yq conda-build jinja2 anaconda-client
 
 /io/conda-build-all -vvv $UPLOAD -- /io/*
 

devtools/osx-build.sh

@@ -10,9 +10,11 @@ brew tap -y caskroom/cask
 curl -s -O https://repo.continuum.io/miniconda/Miniconda3-latest-MacOSX-x86_64.sh;
 bash Miniconda3-latest-MacOSX-x86_64.sh -b -p $HOME/anaconda;
 export PATH=$HOME/anaconda/bin:$PATH;
+conda config --set channel_priority false;
+conda config --add channels conda-forge;
 conda config --add channels omnia;
 conda config --show;
-conda install -yq conda-build=2.1.5 jinja2 anaconda-client;
+conda install -yq conda-build jinja2 anaconda-client;
 
 #export INSTALL_CUDA=`./conda-build-all --dry-run -- openmm`
 export INSTALL_OPENMM_PREREQUISITES=true

ecos/meta.yaml

@@ -8,9 +8,8 @@ source:
   md5: 39cc4c4a5598cc51ca4dc9aec810f41a
 
 build:
-  skip:
-    # ecos/src/ecos.c(1093): error C2065: '_TWO_DIGIT_EXPONENT': undeclared identifier
-    - [win and py35]
+  skip: True # [win and (py35 or py36)]
+  # Disable for now under certain windows conditions
 
 requirements:
   build:

fftw3f/FLAG_FOR_REVIEW_FFTW

@@ -0,0 +1,3 @@
+So this one may have a name colision with FFTW on conda-forge. It appears to be the same thing.
+
+People will probably need to rename their dependencies

openforcefield/meta.yaml

@@ -9,6 +9,7 @@ source:
 build:
   preserve_egg_dir: True
   number: 0
+  skip: True # [win]
 
 requirements:
   build:

openmm-example-plugin/meta.yaml

@@ -8,7 +8,7 @@ source:
   git_rev: f1e12743d7c5403bb6d7ad8da69af6a2f27e1b26
 
 build:
-  skip: True  # [win]
+  skip: True # [win]
 
 requirements:
   build: