This builds Linux and Windows release binaries and sources for NumPy using a Vagrant VM. It should be 100% reproducible. You need Vagrant and Fabric.
The NumPy release notes are here:
https://github.com/numpy/numpy/blob/master/doc/HOWTO_RELEASE.rst.txt
But most of it is automated by numpy-vendor, so once the release branch is approved for release, just follow the instructions below.
Do:
vagrant up fab vagrant prepare
Log in and manually inspect the numpy repository, the release branch will be checked out. Do some last minute fixes, checkout a different branch, etc.:
vagrant ssh cd repos/numpy # Do any changes that are not in official repositories
First, change/check the following variables in pavement.py
depending on the
release version:
RELEASE_NOTES = 'doc/release/1.7.0-notes.rst' LOG_START = 'v1.6.0' LOG_END = 'maintenance/1.7.x'
Do any other changes. When you are ready to release, do the following changes:
diff --git a/setup.py b/setup.py index b1f53e3..8b36dbe 100755 --- a/setup.py +++ b/setup.py @@ -57,7 +57,7 @@ PLATFORMS = ["Windows", "Linux", "Solaris", "Mac OS- MAJOR = 1 MINOR = 7 MICRO = 0 -ISRELEASED = False +ISRELEASED = True VERSION = '%d.%d.%drc1' % (MAJOR, MINOR, MICRO) # Return the git revision as a string
And make sure the VERSION
variable is set properly. Commit it:
git commit -a -m "REL: Release 1.7.0rc1"
Push this commit into your local github account, not the official repository yet, for example by:
vagrah ssh cd repos/numpy git push [email protected]:certik/numpy maintenance/1.7.x:release_test
which will work, because vagrant ssh
forwards the ssh keys.
Do the release (build general and windows binaries from the repos/numpy
directory in the Vagrant VM):
fab vagrant numpy_release
And the directory release
will be created in
the current directory from the VM. If you need anything else, just login using
vagrant ssh
and copy it to /vagrant
inside the VM.
If everything worked correctly, get the release commit from your github, tag it:
git tag v1.7.0rc1
and push this commit and tag into the official repository. Finally, upload the binaries.
If you want to start from scratch, but want to reuse the existing vagrant image, you can keep all the deb packages, but reinstall everything else (NumPy, Wine, ...) from scratch. Do:
fab vagrant remove_userspace fab vagrant prepare_userspace
This will delete all builds and other possible changes that you made into
the vagrant image. This is equivalent to fab vagrant prepare
, but faster.
First prepare the Mac build box as follows:
- Install Python 2.7, 3.3 and 3.4 from python.org using the dmg disk image
- Install setuptools and bdist_mpkg into all these Pythons
- Install Paver into the default Python
Tip: Add the /Library/Frameworks/Python.framework
directory into git and
commit after each installation of any package or Python. That way you can
easily remove temporary installations.
After Linux and Windows release files are in the release
directory, we
generate Mac binaries. First we setup the environment:
fab -H mac_host mac_setup
where mac_host
is the Mac OS X computer which will be used to build the
binaries. The only prerequisite is that the release
directory contains
the reference.pdf
and userguide.pdf
files generated by the Vagrant VM
in the previous step. You can login manually and do some changes in the numpy
directory if needed.
Then generate the binaries by:
fab -H mac_host mac_numpy_release
The Mac binaries will be copied to the release
directory locally.
To remove temporary files from the Mac, use:
fab -H mac_host mac_remove_userspace
Append the checksums at the end of README.txt in the release directory:
md5sum *.dmg >> README.txt
Upload all the files from the release
directory to SourceForge:
http://sourceforge.net/projects/numpy/files/NumPy/
Lookup the last release to see which files need to be uploaded and double check that there are no files missing.