Fix symbolic links to python3 and homeassistant executable. Fix permissions and ownership to homeassistant directory to allow writing to home-assistant.log.
* Pulled through notes of other path options
Added notes that for Hass.io there may be multiple path options depending on the device.
* Updated Hass.io note
* sync usb_path to other docs
The hassio zwave documents reference usb_path: /dev/ttyACM0, so this document should be in sync. After some discussion it appears that this could be variable, in which case there should be a good way to figure out the CORRECT answer rather than using a guess as example code without stating it as such. Possibly linking back to the linux "how to find" section?
* Removed manual set language per suggestion
Raspbian does not keep files in /run. Therefor the pid directory needs to be created every time after boot. The easies is to put this into the start function.
Sorry, I did not test the patch with all my legacy systems.
I changed the init scripts to be more reliable.
hass executable supports the option --log-file. Using this is much safer and does not lead to problems with permission and overwriting old logfiles. Also I added creating the directory and ownership in the install function. To prevent log file groth I added a logrotate rule.
I moved the update-rc.d command you had in your manual into the install function since you already ad it in the uninstall function. This prevents more copy&paste errors.
I changed the PID file to be located in an extra directory. This way you do not have to to this hack with writing a fake PID file in the install function. Also, I read many users having problems that hass daemon is not starting (see eg. https://community.home-assistant.io/t/autostart-virtualenv-ubuntu-14-04/2120/37) without showing any errors. This is due to a missing PID file or its wrong permissions. By changing as stated and adding a test this problem should not arise any more.
I changes some wording in the manual to make it more understandable.
Given that people using Windows keep grabbing the alpha releases of 3.7, I've added a recommendation of Python 3.6, and a note about not downloading alpha or beta versions, and how to identify them.
* Added WallC-S, layout tweak
Moved the central scene configs to their own block, added entry for WallC-S based on feedback in Discord
* 🚑 Fixing the build
* Advise to stop hass before running certbot
Since I had trouble running certbot while home assistant was running, but it worked fine when I stopped home assistant before, I added this to the tutorial.
* Update name
* Update freenas.markdown
The example provided for a chained command to mkdir /home and /home/.homeassistant was missing the second mkdir command. A copy and paste of this example would fail.
* Update freenas.markdown
Updated mkdir commands to use '-p' option.
* Improved zwave security network key description
As someone adding a zwave device securely for the first time I found there were some gaps in the documentation.
* ✏️ Corrected small typo
* Update guide for clarity
Recommends auto-starting by default, with manual starting as an option. Explicitly pull out the upgrade steps (since it also involves stopping HA first). Recommend install as a dedicated user.
* Minor changes
It's often been noted in the forums that Template Triggers are sometimes overload and that you want to use them sparingly. I haven't stress tested my system with this idea, but if the template triggers are tested every time any state changes, then that is the clarification that I've proposed in the documentation.
* Removes extraneous paragraph from docker documentation
It looks like this line was intended to go somewhere else, but got
erroneously bumped into the `Restart` section.
* Adds instructions to expose devices when deploying under docker
* Adds info about `docker-compose.yml` configuration to docker deployments
The docker command is pretty complicated and doesn't support automated
restarts. I think most users with want to use `docker-compose` (or
`docker swarm`) to deploy.
