- These devices sometimes do not respond on the first try
or may be subject to transient broadcast failures, or
overloads. We now try discovery periodically once the
integration has been loaded.
- We used to try this 4x at startup, but that solution
seemed to aggressive as we want to be sure we pickup
the devices after startup as well since the network
will likely be more calm after startup.
* Fixing tplink issues with offline devices during setup.
* Fixing circleci errors.
* Adding code to defer the creation of entities that are not online.
* Addressing code review changes and cleaning up a little.
* Fixing tests and static analysis.
* Adding test to satisfy coverage requirements.
* Resolving merge conflicts.
* Fixing issue where lights don't appear in the integration page.
* Using pyHS100 properties for most sysinfo.
Addressing some PR feedback.
* Addressing some PR feedback.
* Better testing async_add_entities_retry
Testing for static dimmers.
Making greater use of conf constants.
* Fixing all static analysis issues.
* Adding non-blocking call for getting discovering devices.
* Address PR feedback
* {switch,light}.tplink: use deviceid as unique id, fetch name from the device during initialization
* raise PlatformNotReady when no device is available
* Use mac instead of deviceid
* remove name option as obsolete
* Add support for configuration flow / integration
Allows activating automatic discovery of supported devices from the configuration
* Fix linting, update requirements_all.txt
* start cleaning up tplink component based on feedback
* add device info, improve config handling
* Allow overriding detected devices via configuration file
* Update requirements.txt
* Remove debug logging
* make hound happy
* Avoid I/O during init and simplify the code, remove remains of leds_on
* Fix issues based on feedback, use consistent quotation marks for device info
* add async_setup_platform emiting a deprecation warning
* Avoid blocking the I/O, check for None on features
* handle some Martin's comments, schema-validation is still missing
* use async_create_task instead of async_add_job, let core validate the schema
* simplify configuration handling by storing the configuration data separately from initialized instances
* add default values to schema, make hound happy
* with defaults set by schema, simplify the checks. add async_unload_entry
* Use constant for data structure access
* REWORD add a short note about async_unload_entry
* handle feedback from Martin, config_data is checked against Noneness
* use pop to remove the domain on unload
* First steps to add tests for the new tplink component
* embed platforms under the component directory
* Fix tests by mocking the pyhs100 internals
* Fix linting
* Test against multiple instances of devices, tidy up
* (hopefully) final linting round
* Add pyHS100 to test requirements
* log always the warnings occured during an update to make them easy to see
* revert back the warning behavior (requirement for silver level in IQS)
* Unload only when an entry is being loaded and add tests for that
Thanks @MartinHjelmare for pointing this out!
* Fix linting
* Bump the upstream lib, fixes most prominently the HSV setting on bulbs
* Test unloading for all platforms, clear the data storage instead of popping it out, making it possible to reconfigure after removal without restarting hass first
* Use class variables instead of instance variables for bulb states, required for HS220
* Use new-style format string
* Fix indenting, uppercase the mock constant
* Run black on test_init, hopefully that will finally fix the weird formatting (pycharm, pylint and hound seems to have different opinions...)
J. Nick Koston
Teemu R
Michael
Marc Mueller
Franck Nijhof
Paulus Schoutsen
Ville Skyttä
Michał Mrozek
Robert Van Gorkom