Update unifrac-binaries recipe, remove hdf5 dependency

[sfiligoi]

HDF5 is fully statically linked in unifrac-binaries 1.2.
Remove runtime dependencies from the recipe.

---

Please read the guidelines for Bioconda recipes before opening a pull request (PR).

• If this PR adds or updates a recipe, use "Add" or "Update" appropriately as the first word in its title.
• New recipes not directly relevant to the biological sciences need to be submitted to the conda-forge channel instead of Bioconda.
• PRs require reviews prior to being merged. Once your PR is passing tests and ready to be merged, please issue the @BiocondaBot please add label command.
• Please post questions on Gitter or ping @bioconda/core in a comment.

Please use the following BiocondaBot commands:

Everyone has access to the following BiocondaBot commands, which can be given in a comment:

@BiocondaBot please update	Merge the master branch into a PR.
@BiocondaBot please add label	Add the please review & merge label.
@BiocondaBot please fetch artifacts	Post links to CI-built packages/containers. You can use this to test packages locally.

For members of the Bioconda project, the following command is also available:

@BiocondaBot please merge

Upload built packages/containers and merge a PR. Someone must approve a PR first! This reduces CI build time by reusing built artifacts.

Also, the bot watches for comments from non-members that include @bioconda/<team> and will automatically re-post them to notify the addressed <team>.

[sfiligoi]

@BiocondaBot please fetch artifacts

[BiocondaBot]

Package(s) built on Azure are ready for inspection:

Arch	Package	Zip File
linux-64	unifrac-binaries-1.2-h73f19ed_1.tar.bz2	LinuxArtifacts
osx-64	unifrac-binaries-1.2-h6f31009_1.tar.bz2	OSXArtifacts

---

You may also use conda to install these after downloading and extracting the appropriate zip file. From the LinuxArtifacts or OSXArtifacts directories:

conda install -c ./packages <package name>

---

Docker image(s) built (images are in the LinuxArtifacts zip file above):

Package	Tag	Install with docker
unifrac-binaries	1.2--h73f19ed_1	show gzip -dc LinuxArtifacts/images/unifrac-binaries:1.2--h73f19ed_1.tar.gz | docker load

[sfiligoi]

@BiocondaBot please add label

[sfiligoi]

Tested the artifacts locally, and they look good

[bgruening]

The readme has mentioned HDF5 everywhere: https://github.com/biocore/unifrac-binaries#install-native

Is this really how it should be done?

[sfiligoi]

The readme has mentioned HDF5 everywhere: https://github.com/biocore/unifrac-binaries#install-native

Is this really how it should be done?

hdf5 is needed during compile, but not at runtime.
We link against the static hdf5 libraries, so we can co-exist with other HDF5 library versions (if any) without any problems.

PS: We could possibly improve the README, yes.

[ebolyen]

@sfiligoi, but that wasn't previously the case correct? It looks like that was added/fixed in biocore/unifrac-binaries#15, which is included in this latest release.

(Thanks for updating unifrac by the way! We had just tracked down some incompatibilities between h5py/scikit-bio and unifrac-binaries in our latest pre-release envs, and these changes should fix that)

[sfiligoi]

@ebolyen Yes, unifrac-binaries 1.1.x needed hdf5 at runtime.

Note: The hdf5 dependency in 1.1.x was actually a Makefile bug; the dependency was never supposed to be there.
But that was fixed in biocore/unifrac-binaries#15

[ebolyen]

Anything we can do to help get this merged?

(I realize this is isn't ideal, but we're waiting on these changes so that we can get the latest scikit-bio in the next QIIME 2 release.)

[ebolyen]

We've tested this locally as well now and can confirm that the static linking appears to be working.

We succesfully installed these built packages alongside a source install of unifrac with the latest version of hdf5 (1.12.2) which differs from what unifrac-binaries was statically-linked against (1.12.1).

The tests from our downstream packages (q2-diversity-lib) also succeed with these new versions.

[ebolyen]

(I'm not really sure why bioconda built with 1.12.1 and not 1.12.2, but it's what appears in the rendered recipe inside the package.)

[sfiligoi]

@bgruening Anything blocking this revision?

[ebolyen]

Is anyone on the Bioconda team available to review this? Thanks!

[sfiligoi]

@rpetit3 Any chance you can look at this?

[sfiligoi]

@matthdsm Any chance you can review this?

[matthdsm]

@BiocondaBot please merge

[BiocondaBot]

I will attempt to upload artifacts and merge this PR. This may take some time, please have patience.