>Hi,
>at first I could not follow your arguments. The it became clearer.
>However, I think, it is true only if the device connected between A and B
>has enough capacity to hold some coulombs during pin transition. On the
>other hand:

That's correct, but I don't know how to determine what is going to be too
much.
I don't know if microchip does either.  The capacitor presented by a piezo
beeper is fairly large in these terms, so it's a significant hazard. Not for
the
casual experimenter, but if you make 1000 of your device, maybe some
number will suffer unexplained crashes, maybe none, maybe all.


>the piezo devices drain not too much current, i. e. you can put a serial
>resistor between it and the PIC. Now, assuming the capacity is high enough
>to satisfy the condition described above, nothing bad will happen as the
>PIC has voltage clamping diodes. I guess 2 x 10k will work.


Not that large an R.. Maybe 100 ohms or so, but 10k won't leave you any
sound.
It's not a problem at all if you sequence the outputs properly.