This increases the global limit to 6535 seconds (1h48m55s) because that is
supported by all light platforms.
Some platforms support even longer transition times so the limit should
actually be platform specific.
* Allow configurable conditions for Pi-Hole sensor
* Include all three conditions by default
* Share Pi-Hole API data across all sensors; eliminate redundant API calls
* Add effect support to MQTT Light
* Use effect state topic for supported_features
* Dont use rainbow as default color
* Add color_temp support to MQTT JSON Light
* Add effect to MQTT JSON light
* Support lights in MQTT discovery
* Allow discovered devices to set their platform
* Add white value support to MQTT Light
* Add white value support to MQTT JSON Light
* Remove blank line
* Add color_temp support to MQTT Template light
* Add white value support to MQTT Template Light
* Remove unused SUPPORT_MQTT_TEMPLATE and stale unused flash and transition code from MQTT Template
* Add XY Color to MQTT Light Platform
* Fix syntax
* Fix more syntax errors
* Revert "Remove unused SUPPORT_MQTT_TEMPLATE and stale unused flash and transition code from MQTT Template"
This reverts commit c03798cb63.
* MQTT Template supports flash and transition but doesnt allow templating of the values
* Add XY color support to MQTT JSON
* Proper variable names
* Only allow whitelisted MQTT platforms to be loaded via MQTT Discovery
* Minor tweaks.
* Add support for remove services / Reload script support
* Reload support for scripts
* Add more unittest for services
* Add unittest for script reload
* Address paulus comments
* Passing pep8, no tests yet
* Fixed some issues with the request throttling
* Removed ability to set throttle time because it was causing more issues than it was worth
* Added blink to .coveragerc
* Changed blinkpy version
* Removed global var, fixed per PR requests
* Added services for camera, migrated switch to binary_sensor
* Added schema for service, fixed naming, removed unused function
Current implementation of connection status doesn't follow convention and is not properly configurable. Might be added again in the future as a full fledged entity or some other way.
For now users can rely on error logging to determine connection status.
* Update to pylgtv 0.1.4
* Send icon with webostv notifications
Default to the homeassistant logo, but allow customizing it on the
component and for individual notifications
* Wait up to 9 seconds
* Set number of recorder retries to 8
* Do not sleep when reporting last connection error if no retries left
* Make sure we clean up old engine if connection is retrying
* Update __init__.py
* Use bundled certificates if port matches mqtts
* Move import requests.certs to top, since it's used in more places
* Add happy and non-happy path tests for default certificate bundle on mqtts port
* Add Zwave refresh services
* services file
* Use dispatcher
* Add zwave prefix to signal
* Delay zwave updates for 100ms to group them.
* Fixes
* lint
* Access _scheduled_update from loop thread only.
* More async
* Some optimizations
* Fix
Message deletion in IMAP is a two step process: first delete, then expunge.
Deleting a message just sets a flag that usually makes the mail client hide
the message. It is the expunge that actually removes the message.
Thus, exclude the deleted messages so that the unread count matches up with
that of most mail clients.
A LIFX bulb maintains its previous color even when the light is off.
For example, if the previous color is blue and the bulb is turned on
and then set to a red color, it will transition through purple colors.
After this commit, the target color is set while the bulb is still
turned off. This overrides the previous color and brightness that the
bulb remembered. The light is then turned on with the requested
transition duration.
For the example, this gives the expected result of only going through
red colors.
* Add ComEd RRTP price sensor
* Update wording to reflect ComEd's naming change from 'RRTP' to 'Hourly Pricing'
* Changed name of sensor source file
* Cleanup based on requested changes
* More cleanup
* small cleanups
* Don't initialize components which have already been discovered (fixes#5588)
* Don't log that we've found a service unless we know it's not a duplicate
* Encode discovery data hash with JSON
This also solves the issue of trying to hash non-hashable objects like dicts
* Add test for duplicate device discovery
* Added support for multiple codes executed in a row
now codes can be specified either by simply providing a single code, which will then be sent like usual, or multiple codes can be executed in a row, specified in a comma delimited format in the configuration.yaml. For example: 111111,222222,333333,444444 would mean 111111 would be sent first, followed by 222222 and 333333 and 444444.
* rpi_rf: added line breaks to not exceed 79 characters per line
* include validation for correct formatting of codes
added regex which only allows either a single number (like 1252456245) or a sequence of commas followed by another number.
* added line breaks to not exceed 79 characters per line
* fix for 'continuation line under-indented for visual indent'
* another try at 'continuation line under-indented for visual indent'
* changed from regex to list for easier maintainability
* removed unnecessary splitting of strings
* Integrate suggestion in #5590 by nordlead2005. This change has been
sitting in limbo for over a month, but it is a good idea. I don't
mean to step on nordlead2005's toes, but we need to make progress.
* Use defined constant for TEMPERATURE_HOLD
* Integrate handling of vacation into hold mode. Canceling vacation
hold requires an update to the external pyecobee library. Creation
of vacation is not supported (it would be straightforward in the code,
but a complex user interface would be required, similar to what is
now done in the ecobee thermostat).
* Add capability to retrieve list of defined climates from ecobee.
* The mode() method used to return the system mode in internal
representation. However, the user sees a different notation in
the ecobee thermostat. Seeing some internal name is particularly
weired with user-defined climates, where these are named "smart1",
"smart2", etc., instead of the name the user has defined. Return
the user-defined name instead. This change might break some user
interfaces but is easily remedied (e.g., use "Away" instead of
"away").
* Simplify is_away_mode_on().
* Correction of erroneously indented else statement.
* Change comment as flake8 gets confused.
* Make zwave devices listen on less network changes.
* Convert more platforms
* Remove printouts.
* Fix copy-paste
* Change default dependent list to empty list
* Fix possible race at startup in mysensors callback
* Update devices via persistence before starting gateway to avoid
two threads calling the same callback at the same time.
* Call add_devices max once per callback
There is no reason for the H to be capitalized. Changing it to lowercase increases consistency with other components and allows for use of the min/max sensor, which throws an error if the units of measurement are not the same.
* added frontier_silicon constant
* added the frontier_silicon component
* cleaning up according to travis
* trying to satisfy pylint
* trying to satisfy pylint
* fsapi version 0.0.6
* with fsapi version 0.0.7
* added fsapi dependency
* yielding the FSAPI
* Removing white space from docstring
* Removing white space from an empty line
* Switching to sync
* clean up white spaces and rename device to FSAPIDevice
* added frontier_silicon constant
* added the frontier_silicon component
* cleaning up according to travis
* trying to satisfy pylint
* trying to satisfy pylint
* fsapi version 0.0.6
* with fsapi version 0.0.7
* added fsapi dependency
* yielding the FSAPI
* Removing white space from docstring
* Removing white space from an empty line
* Switching to sync
* clean up white spaces and rename device to FSAPIDevice
* changed info to debug
* added frontier_silicon constant
* added the frontier_silicon component
* cleaning up according to travis
* trying to satisfy pylint
* trying to satisfy pylint
* fsapi version 0.0.6
* with fsapi version 0.0.7
* added fsapi dependency
* yielding the FSAPI
* Removing white space from docstring
* Removing white space from an empty line
* Switching to sync
* clean up white spaces and rename device to FSAPIDevice
* added the frontier_silicon component
* trying to satisfy pylint
* fsapi version 0.0.6
* remove white space
* generated requirements
* added the frontier_silicon component
* cleaning up according to travis
* trying to satisfy pylint
* trying to satisfy pylint
* fsapi version 0.0.6
* with fsapi version 0.0.7
* added fsapi dependency
* yielding the FSAPI
* Removing white space from docstring
* Removing white space from an empty line
* Switching to sync
* clean up white spaces and rename device to FSAPIDevice
* trying to satisfy pylint
* changed info to debug
* added the frontier_silicon component
* fsapi version 0.0.6
* generated requirements
* pylint
* moved import requests to the method where it is being used
* add a basic unit test
* cleaned up source code
* added frontier_silicon constant
* added the frontier_silicon component
* added basic test
* added fsapi to requirements_all.txt
* added coverage omit, though a basic test was included
* added MEDIA_TYPE_MUSIC for artist and album
* removed duplicate cons
* switched fsapi call to a property, removed unecessary comment
* detailed docstring for fs_device
* added a space for the info_name - info_text separator
* reduced proeprty (fsapi) access for volume down/up
* Fix: replace influxdb query by another query that is more lightweight and won't timeout
* Fix: replace influxdb query by another query that is more lightweight and won't timeout
* Fix for OSRAM lights connected to hue bridge
Do not send command "effect = none" to OSRAM lights
Osram lights connected to a hue bridge do not seem to handle "effect =
none" very well. Most of the times they jump to the selected color and
then change to red within a second.
Osram lights connected to a hue bridge do not handle xy values outside
of their gamut. Since they just stay at their old color value, handling
the UI is very unpredictable. Sending HSV values to the lights fixes this.
* Add tests for new util methods
* Fast & efficient updates for ZAMG weather data
ZAMG updates on the hour, so instead of checking every half-hour we can
check each minute - only after the observations are taken until
receiving them.
* sensor.zamg: test instead of whitelist for station_id
* Autodetect closest ZAMG station if not given
* ZAMG weather component, based on the sensor
* Review improvements
* Update to new ZAMG schema, add logging
Turns out it wasn't a typo, but rather an upstream schema change. Added
better error handling to ease diagnosis in case it happens again.
* No hardcoded name
- Separate attribs from coffeemaker condition
- Set power units for threshold to mW to be consistent with others
- Adjust on-time labels to be more clear
* Create zwave devices on OZW thread and only add them during discovery.
* Read and write devices dict from loop thread.
* More async
* replace callback with coroutine
* import common function instead of callin git
* Bump qnapstats library to 0.2.3
* Expose the new timeout setting
* Show persistent notification if QNAP fails to set up
This sitaution will usually occur if the server configuration is wrong or the
timeout is set too low. In both cases `api.update()` will fail, which is a
problem because we cannot initialize many of the sensors without having this
data.
* Add new system_temp condition to QNAP sensor
Avoid unhandled exception and stack trace when server closes connection by changing from handle ClientDisconnectedError to DisconnectedError. Also added HttpProcessingError, which was missing.
* Connect to PLM and process simple protocol callbacks
* Baseline commit
* Connect to PLM and process simple protocol callbacks
* Baseline commit
* Connection working again
* Async add devices is working via callback now
* Beginning to interface with PLM library for control and state
* Deal with brightness in 255 levels with library
* Change sub names to match API changes
* Remove PLM-level update callback
* Support dimmable based on underlying PLM device attributes
* Expand to non-light platforms
* Stubs for turn on and off
* Current version of Python library
* Amend to use switch device attributes
* Use asyncio endpoints for control
* Add logging line
* Bump module version to 0.7.1
* Auto-load platforms, display device info/attributes
* Unify method name for getting a device attribute
* Require Current version of insteonplm module
* Import the component function in each platform in the balloob-recommend manner
* For consistency, handle switch state as onlevel just like lights
* Use level 0xff for on state, even with binary switches
Observing the behavior of a 2477S switch, it looks like even the non-dimmable
devices use 0x00 and 0xff for off/on respectively. I was using 0x01 for on
previously, but that yields unnecessary state change callbacks when message
traffic ends up flipping the onlevel from 0xff to 0x01 or 0x01 to 0xff.
* Use sensorstate attribute for sensor onoff
* Move new device callback to devices attribute
* Add support for platform override on a device
* Bump version of insteonplm module
* Default overrides is an empty list
* Avoid calling private methods when doing common attributes
* Remove unused CONF_DEBUG for now
* flake8 and pylint code cleanup
* Move get_component to local function where it is needed
* Update to include insteonplm module.
* New files for insteon_plm component
* Legitimate class doctring instead of stub
* Docstring changes.
* Style changes as requested by @SEJeff
* Changes requested by @pvizeli
* Add @callback decorator to callback functions
* Opportunistic platform loading triggered by qualifying device detection
Instead of loading all the constituent platforms that comprise the insteon_plm
component, instead we defer and wait until we receive a callback for a device
that requires the platform.
* Restore states
* feedback
* Remove component move into recorder
* space
* helper
* Address my own comments
* Improve test coverage
* Add test for light restore state
* Set configurable refresh value delay to 5 seconds
* Revert "Remove configurable refresh value delay, and fix it to 5 seconds"
This reverts commit edc2dc35d165e5e13b0f3cf4df40493a7ce764f1.
* Use default value of 5 sec for refresh, and still configurable
* Make default delay 5secs, but needs to be activated
* Added support for alternate SSH ports in AsusWRT (#4832)
* Always set the SSH port in SSH arguments
* Removed whitespace
* Adding port=22 to the mock calls
* Changed config.get(CONF_PORT) to config[CONF_PORT] per feedback from @pvizeli
* Added forecast functionality to the weather platform, updated OWM to get forecast data
* style fixes
* Changed returned forecast data to a more managable dict
* Fixed line length
* forecast will always be collected, data returned from owm is based on metric setting
* use list comprehension to create the forecast data
* Added forecast data to the weather demo
* Create dict directly in list comprehension
* Minor variable change in weather demo platform
* Convert forecast temperatures in weather component
* set forecast attributes as constants
* Style fixes and tests
* Copied forecast_entry instead of mutating data
* Support aliased owfs sensors
Current implementation does not support OneWire OWFS sensors that are aliased by OWFS alias.txt file:
http://owfs.org/index.php?page=aliases
Assumption is that folder name == sensor address
This change reads the supported families from owfs "family" file and adds both aliased and unaliased sensors.
(this approach also skips the "management" folders.. eg simultaenous, bus, alarm, statistics etc. For example when adding "simultaneous" as sensor (it also has "temperature" file) then owfs crashes)
* Update onewire.py
lint suggested style fixes
* Update onewire.py
Empty line removed
* Update onewire.py
comments removed
* Migrate mqtt to async
* address paulus comment / convert it complet async
* adress paulus comment / remove future
* Automation triggers should be async
* Fix MQTT async calls
* Show that event helpers are callbacks
* Fix tests
* Lint
* First round of not using values directly
* Round two
* lint
* Round four
* Conflict
* Round five
* Update zwave.py
* round six
* Docstring
* flakywakie
* Fetch values in constructor
* Blank line removal
* Set attributes in callback
* Docstring
* Round seven
* Ughgit add homeassistant/components/lock/zwave.py!
* Sloppy code
* Add aurora sensor
* allow custom forecast threshold for aurora binary sensor
* move AuroraGateway functionality to Aurora data object to conform with HA standards
* Add fake support for turn on/off for Apple TV
When the device is "turned off", no requests are sent to the device.
When the setting "start_off" is set to true, the device starts in off
state.
* Fix async comments
* Clean up supported features
* Reconnect robustness, expose connection state.
- Expose connection status as rflink.connection_status state.
- Handle alternative timeout scenario.
- Explicitly set a timeout for connections.
- Error when trying to send commands if disconnected.
- Do not block component setup on gateway connection.
* Don't use coroutine where none is needed.
* Test disconnected behaviour.
* Use proper conventions for task creation.
* Possibly fix test race condition?
* Update hass import style
* Add support for the Open Energy Monitor Thermostat
* Fix linting errors
* Define an update method and local state
* fix linter
* Small tweaks.
Update oemthermostat version, default name and docstrings
* Fail to setup oem platform if connection fails.
* update requirements
* More tweaks to auth and exceptions
* Remove target temp
* Bump pwaqi to 1.4 to fix a typo in the underlying library.
* Update WAQI sensor to use pwaqi 2.0 which relies on AQICN public API.
This is a breaking change as the component now requires 'token' parameter.
* Fix lint
Robbie Trencheny
Adam Mills
Christiaan Blom
Anders Melchiorsen
happyleaves
Pascal Vizeli
pvizeli
Paulus Schoutsen
ArrayLabs
Reed Riley
Colin O'Dell
siebert
Kevin Fronczak
Johan Bloemberg
Kevin Siml
Josh Anderson
Barry Williams
Markus Peter
Dennis de Greef
Andrey
Job Vermeulen
Igor Shults
Thibault Cohen
Marcelo Moreira de Mello
Teemu R
Johann Kellerman
Daniel Høyer Iversen
Lev Aronsky
joe248
John Mihalic
Valentin Alexeev
Open Home Automation
Micha LaQua
Rowan
Jan Losinski
Alan Fischer
Pierre Ståhl
martinfrancois
Duoxilian
Mitko Masarliev
Fabian Affolter
Martin Hjelmare
Erik Eriksson
Stefano Scipioni
jumpkick
ericgingras
Krasimir Zhelev
Boris K
arjenfvellinga
Lindsay Ward
Greg Dowling
Daniel Perna
pavoni
TimV
Wolf-Bastian Pöttner
groth-its
Nate
Jose Juan Montes
Jeff Wilson
Scott Henning
Philipp Schmitt
Erik
Zac Hatfield Dodds
Justin Dray
David McNett
Matt N
Lewis Juggins
John Arild Berentsen
Scott Bradshaw
Tom Matheussen
normakm
William Scanlon
dramamoose
kitcorey
Walker Boyle
Jon Caruana
George.M
Fabian Affolter
Tom Dickman
Sören Oldag
PetePriority
Gert-Jan van de Streek
Stuart Mumford