Yes, I think we have to repeat it over and over and over a gain, as we have to tell a little child: "No!! Don't touch the oven!! It's hot!! It will hurt you!!"

The best argument against simple RC-reset circuits is, that it can cause a loss of code lines stored in the internal flash code memory. This was reported many times in application notes, erratas and other publishings. This is reported even today, and this will be reported in the future.

Do never use nonreliable reset "solutions" in combination with a flash micro!

People often forget, that the micro is a rather dynamically working machine. Many micro-routines are running, whenever we think that a rather static procedure is in process. But that's wrong. That's also the reason, why even small glitches on port lines or Vcc line can cause the micro to run ill. Because then a complicate internal micro-routine was given an unsane "kick" or was even "invoked", like the "erase flash" or "write to flash" micro-routines!!

Some people might tell me now: Yes, I have noticed that a toy hangs-up from time to time and that I have to switch it off and and on again. But I never have noticed a loss of code lines. So, what you tell me can't be valid!

Well, toys are seldom made using flash micros, but OTP-Eproms or even ROMs, for cost reasons. Would flash micros be used, you would have noticed the loss of code lines...

Kai