* Fixed the issues related to auto mode
I was having the same issues as described in #63403, specifically, the error stating that Mode 7 is not valid, only Heat, Cool, Off when trying to do anything while the thermostat is set to Auto. This error originates with the way the Lyric API handles the modes. Basically, when one queries the changeableValues dict, you get a mode=Auto, as well as a heatCoolMode, which is set to either Heat, Cool, Off. Per the documentation, heatCoolMode contains the "heat cool mode when system switch is in Auto mode". It would make sense that when changing the thermostat settings, mode=Auto should be valid, but it's not. The way the API understands that the mode should be set to Auto when changing the thermostat settings is by setting the autoChangeoverActive variable to true, not the mode itself. This require changes in the async_set_hvac_mode, async_set_temperature, and async_set_preset_mode functions. Related to this issue, I got rid of the references to hasDualSetpointStatus, as it seems that it always remains false in the API, even when the mode is set to auto, so again, the key variable for this is autoChangeoverActive.
While I was working on this I also noticed another issue. The support flag SUPPORT_TARGET_TEMPERATURE_RANGE had not been included, which did not allow for the temperature range to be available, thus invalidating the target_temperature_low and target_temperature_high functions. I added this flag and sorted out which set point (heat vs cool) should be called for each of them so things work as expected in Lovelace. I have tested all of these functionalities and they all work great on my end, so I thought I'd share.
* Update climate.py
* Update climate.py
Fixed two additional issues: 1) When the system is turned off from Auto, the heatCoolMode variable becomes 'Off', so when you try to restart the system back to Auto, nothing happens. 2) I now prevent the async_set_temperature function from being called with a new set point when the system is Off.
All changes tested and functional.
* Update climate.py
* Update climate.py
Return SUPPORT_PRESET_MODE flag only for LCC models (i.e. they have the "thermostatSetpointStatus" variable defined). TCC models do not support this feature
* Update climate.py
After playing with the official Honeywell API, I realized it doesn't like to received commands with missing data, i.e., it always wants to get a mode, coolSetpoint, heatSetpoint, and autoChangeoverActive variables. This was causing some random issues with changing modes, especially from coming from off, so I modified the async_set_temperature, and async_set_hvac_mode fuctions to always send all pertinent variables.
* Update climate.py
* Update climate.py
* Update climate.py
* Update climate.py
* Clean code and test everything
Alright, sorry for the multiple commits, fixing this properly took a fair bit of testing. I went ahead and cleaned up the code and made the following big picture changes:
1) The integration now supports the Auto mode appropriately, to include the temperature range.
2) There's a bug that actually manifests when using the native app. When the system is 'Off' and you try to turn it on to 'Auto', it will turn on briefly but will go back to 'Off' after a few seconds. When checking the web api, this appears to be related to the fact that the heatCoolMode variable seems to continue to store 'Off', even if the mode accurately displays 'Auto', and the autoChangeoverActive=True. So to overcome that inherent limitation, when the system is 'Off' and the user turns it to 'Auto', I first turn it to Heat, wait 3 seconds, and then turn it to 'Auto', which seems to work well.
* Update climate.py
* Fixed errors
* Fixed comments that were resulting in error.
* Update climate.py
* Update homeassistant/components/lyric/climate.py
Co-authored-by: Martin Hjelmare <marhje52@gmail.com>
* Update homeassistant/components/lyric/climate.py
Co-authored-by: Martin Hjelmare <marhje52@gmail.com>
* Update climate.py
I removed a blank line in 268 and another one at the end of the document. I also fixed the outdents of await commands after the _LOGGER.error calls, not sure what else may be driving the flake8 and black errors. Any guidance is much appreciated @MartinHjelmare
* Update climate.py
* Update climate.py
corrected some indents that I think were the culprit of the flake8 errors
* Update climate.py
I used VS Code to fix locate the flake8 errors. I ran black on it, so I'm hoping that will fix the last lingering black error.
Co-authored-by: Martin Hjelmare <marhje52@gmail.com>
* Change debug level of integration reload after config update to debug
* Remove extra .keys() call for dict itreation
* Remove unecessary type check
* Remove unused LOGGER reference
The base `Cover` entity requires an explicit value for
`_attr_is_closed`.
Since the `SomfyShade` is an assumed state, we don't know
by default whether the shade is open or not, so we need to
explicitly return `None` for `_attr_is_closed`
* Align selectors with frontend updates
* Drop metadata from MediaSelector selection
* Adjust blueprint tests
* Address review comments
* Add tests for new selectors
* Don't stringify input
* Require min+max for number selector in slider mode
* vol.Schema does not like static methods
* Tweak
* Small performance tweaks to _async_write_ha_state
- Only call self.available once per cycle
- Only call self.device_state_attributes once per update cycle
- Do not check for device_state_attributes if extra_state_attributes is not None
* drop backcompat
* remove prop
* Make google calendar loading API centric, not loading from yaml
Update the behavior for google calendar to focus on loading calendars based on the
API and using the yaml configuration to override behavior. The old behavior was
to first load from yaml, then also load from the API, which is atypical.
This is pulled out from a larger change to rewrite calendar using async and config
flows.
Tests needed to be updated to reflect the new API centric behavior, and changing
the API call ordering required changing tests that exercise failures.
* Update to use async_fire_time_changed to invoke updates
* Mqtt Notify service draft
* fix updates
* Remove TARGET config parameter
* do not use protected attributes
* complete tests
* device support for auto discovery
* Add targets attribute and support for data param
* Add tests and resolve naming issues
* CONF_COMMAND_TEMPLATE from .const
* Use mqtt as default service name
* make sure service has a unique name
* pylint error
* fix type error
* Conditional device removal and test
* Improve tests
* update description has_notify_services()
* Use TypedDict for service config
* casting- fix discovery - hass.data
* cleanup
* move MqttNotificationConfig after the schemas
* fix has_notify_services
* do not test log for reg update
* Improve casting types
* Simplify obtaining the device_id
Co-authored-by: Erik Montnemery <erik@montnemery.com>
* await not needed
Co-authored-by: Erik Montnemery <erik@montnemery.com>
* Improve casting types and naming
* cleanup_device_registry signature change and black
* remove not needed condition
Co-authored-by: Erik Montnemery <erik@montnemery.com>
Erik Montnemery
Marc Mueller
Franck Nijhof
Numa Perez
Robert Svensson
Tom
Per Öberg
epenet
Allen Porter
Benjamin
J. Nick Koston
Paulus Schoutsen
Gido
Shay Levy
Sean Vig
Jan Bouwhuis
Jeef
G Johansson
GitHub Action
Dmitry Kosenkov
Mike Fugate
Frank
Christopher Thornton
Brandon Rothweiler
Steve Easley
Jesse Hills
jan iversen
Davor Val Vega
Hmmbob
Joakim Sørensen
Robert Hillis
Joakim Plate
starkillerOG
jjlawren
Diogo Gomes
Guido Schmitz
Artem Draft
Thomas Schamm
Tom Harris
Kevin Stillhammer
Álvaro Fernández Rojas
escoand
Raman Gupta
Poltorak Serguei
cheng2wei
Richard de Boer
uvjustin
Lester Lo
Aaron Godfrey