MX682X

Based on the fact that a test with my suggested changes did not generate a single error in 16 hours, I don't think slaveTrasactionOpen() will require the use of sei()/cli()...

> was the compiler inlining or calling the slaveTransactionOpen() I thought yours inlined because you compiled 1xWire and mine did a call because I compiled 2xWire. Just double checked, both...

You can also find the updated Wire library in my fork of DxCore

> open trashes all future use of the slave until RST This is because the slave keeps sleeping. It should be enough to wake the slave up. The reason why...

I've started with debugging right now and I stumbled upon 3 things so far that might create problems: 1. Do you make sure that Serial finishes before putting the CPU...

Since I sometimes fail to convey my thoughts properly because I might leave out some knowledge I assume other people already have: For clarification: a TWI transaction consists of ```...

[twi.c.pdf](https://github.com/SpenceKonde/DxCore/files/9331203/twi.c.pdf) (too lazy to archive, just remove the .pdf after downloading) @ObviousInRetrospect Try to replace your twi.c with this one. It seems to work on my end, at least it...

sender: attiny1614 with modded TWI buffer size: ``` #include "Wire.h" void setup() { pinMode(PIN_PA3, INPUT_PULLUP); Wire.usePullups(); Wire.swap(0); Wire.begin(); } void loop() { if (digitalRead(PIN_PA3) == LOW) { Wire.beginTransmission('A'); for (uint8_t...

> make it error out if you call enableDualMode and have set master or slave in the menu This would be a misleading error, since you can use the dual...

Alright, so seems like I've introduced a bug that automatically called the onReceive function even though no data was available. I should add that I wanted to go step by...