Hi Stew,
it works flawlessly!
I have changed the +8 from one pointer to the other, it was a mistake of course.
I have been forced to save the contents of $ff01-03-21-23 to ensure everything
'still' works alright when the stackblast ends ans returns to the main loop
There is no sign of the copy on the screen, I mean that it is unnoticiable, great!
On the other side, I admit that loading right to $c00 gives roughly the very same result
on an emulated Dragon32.
You have shown us a very nice trick that allows more work to be done between two FS
than one could have imagined. CycleCounting was not enough to see it was doable!
cheers
pere