Open tomjn opened 1 year ago
Set up several shell scripts, they aren't finished though
In the hyper-v provisioner scripts, they have special shared commands that output banners for the start and end of messages intended for vagrant. Then, the Hyper-V provider uses regex to monitor the stdout
to extract error and success messages.
For now we'll use the return code before exploring more in depth options. It may be the odd case where we need to ensure the output of the script is just a process ID or file path that can be checked for by vagrant and read in
Tart have said they don't know much Ruby and can't help with that, which is fine, it's not my strongest language either.
Hashicorp chose to implement their Hyper-V provider in a way that calls a scripts folder with powershell commands. I imagine this is because they couldn't run those commands directly in Ruby. It does mean that someone who knows powershell can contribute, and all the Hyper-V interaction specific parts are neatly arranged in standalone files.
I'm choosing to do the same here. Why faff around with executing things and watching in Ruby if I know how to do it in bash already? Most of the people I'm working with know far more shell script than Ruby/Vagrant APIs.