On Sat, Apr 18, 2015, at 02:22 AM, James wrote: > I just noticed: > https://lists.projectatomic.io/projectatomic-archives/atomic-devel/2015-April/msg00027.html > > I've just posted a message to this list about the oh-my-vagrant > patches that I just wrote. For whatever reason, the archives don't > seem to update very quickly, so I can't link you to the message! > > In any case, I think my patches are feature complete with > https://github.com/cgwalters/vagrant-atomic-cluster (I just had a > quick look) but also add much more functionality, in particular > because you get the existing oh-my-vagrant features too.
This is basically https://github.com/purpleidea/oh-my-vagrant/commit/1f26e5bda5d7585f7b26babcb56a529cc34b96bd right? The thing I really like about Vagrant is the ease of use for linking in *real* config/management/scrripting tools (Ansible/Puppet) etc. (I personally think Vagrant's built-in provisioning like the Docker provisioner are mostly lame hacks) The idea that you can easily test Ansible code locally in a Vagrant cluster on your laptop, and then be able to run those same scripts against a test cluster in AWS/GCE/etc. is powerful. But because you're not using kubernetes-ansible here that means we have less of a shared base. Maybe that's not an immediate burning problem...but from my perspective you're not actually obsoleting cgwalters/vagrant-atomic-cluster because I still want to hack on kubernetes-ansible. Does that make sense?