Fix incorrect subtraction of 2 uint16_t integers in
various examples and README.md docs for 32-bit processors.
On 8-bit processors with sizeof(int) == 2, a uint16_t - uint16_t
returns a uint16_t.
But on 32-bit processors with sizeof(int) == 4, a uint16_t - uint16_t returns an int, which returns a negative number if the
unsigned integer rolled over.
The solution is to explicitly add a (uint16_t) cast, which is
harmless on 8-bit processors, but is required to produce the correct
value on 32-bit processors.
Upgrade various tool chains:
Arduino IDE from 1.8.13 to 1.8.19
Arduino CLI to 0.20.2
Arduino AVR from 1.8.3 to 1.8.4
STM32duino from 2.0.0 to 2.2.0
ESP8266 from 2.7.4 to 3.0.2
ESP32 from 1.0.6 to 2.0.2
Teensyduino from 1.53 to 1.56
No signficant changes except for LadderButtonConfig on ESP8266
(faster) and ESP32 (slower).
Downgrade support for SAMD21 boards to new "Tier 3: May work but
unsupported".
I can no longer flash my SAMD21 boards as of Arduino IDE 1.8.19.
This is a maintenance release. No functional change.
uint16_tintegers in various examples and README.md docs for 32-bit processors.sizeof(int) == 2, auint16_t - uint16_treturns auint16_t.sizeof(int) == 4, auint16_t - uint16_treturns anint, which returns a negative number if the unsigned integer rolled over.(uint16_t)cast, which is harmless on 8-bit processors, but is required to produce the correct value on 32-bit processors.LadderButtonConfigon ESP8266 (faster) and ESP32 (slower).