peterhinch
commented
1 year ago Apologies. I think it took me a while to grasp that the problem arises if further edges occur during the latency period between the first edge and the ISR actually running. Encoders are much more tricky than people realise...
I have added a credit for your contribution to encoder.py.
Hi Peter,
just to give a feedback now: I checked your Encoder code and it looks fine. After introducing the
if (x := pin_x()) != self._x:the flawed behaviour that I once mentioned (a year ago) should be gone. I tried to point you to the flaw then, without success. Later I saw that you had corrected the code without giving credit to my discussion and was disappointed. Later again I saw you got the point by another issue mentioned later. :-)