Thanks. Yes, it's definitely related to the overlap.I don't think you are doing anything wrong. This looks to be a bug in beebasm.py, line 107.
Once the copyblock is done, we need to be smarter about the clear command.
The ranges of memory involved: [from, from+length) and [to, to + length) can overlap.
We should be working out how much of these ranges overlap and only blank out the bit that doesn't overlap.
Since an overlapping move is quite common (particularly when moving games down to &e00), I had assumed someone would have tried this previously. This is why I thought I was doing something wrong. Obviously not!
Anyway, I can work around it by manually tweaking the beebasm file, once it's been generated by py8dis, so no big deal.
Statistics: Posted by KenLowe — Sun Jan 05, 2025 11:05 pm