Dave Jones

CompositeDevice would be the logical ancestor at the moment, but I'm actually re-working that bit of the hierarchy right now after @lurch (rightly) pointed out that AnalogInputDevice really isn't a...

Personally I'm not too bothered about the 0/1/False/True thing since as you note they automatically coerce (right down to the point that you can do things like indexing an array...

I think there's some misapprehension here: "+1" and "-1" are the final states in the state-machine that the encoder moves through with each notch (this is illustrated in [the comments...

@JohanAR in your case I'm guessing (from a quick skim of the code, and the fact the class is called `Knob` :) that you're building something for a rotary encoder....

> Following on from #971 I've just read the [RotaryEncoder](https://github.com/gpiozero/gpiozero/blob/master/gpiozero/input_devices.py#L982) code for the first time. I thought you might like some feedback... ping @waveform80 Always welcome! > The `threshold_steps` feels...

Lots of good ideas here for RotaryEncoder in 2.0. I'll have a play with some of these and see how things "feel". Targetting for 2.0. > As far as I...

I think I see the issue here. You must have an edge coming in precisely during construction of the `Button`. The initial call to `_fire_events` in `InputDevices.__init__` *assumes* that it's...

The docker info *might* be useful to add to the docs (I'm skeptical, but hey - it can't harm), but I have to say that using gpiozero just to get...

Yeah - it opens up a lot of possibilities as there's plenty of HATs out there with SPI and I2C

Your second script (ending with the bare `led.blink()`) will terminate immediately when run from the command line because there's nothing following the blink call. The script ending will shut down...