And now actually the 4th attempt!!
> 3rd attempt...
> 
> I was able to verify by writing some garbage on the screen that the
> booter starts to copy the boot code into low RAM, but never completes.
> If I only copy a small part of the buffer (say, 800 bytes) it will
> get to the next line (which is where I copy junk into the vram).
> If I attempt to write the whole buffer, however, it never finishes the
> copy operation.
> 
> I will be narrowing down the exact limit ofhow many bytes may be copied
> unless I can getsome advice to do something more fruitful.  Like maybe
> putting an assembly code program into the buffer instead of the BSD
> kernel that does some externally visible stuff?
> 
> My best guess for now is that I am overwriting my own copycode routine.
> 
> I tried skipping the copy and jumping straight into the code (jump_point=
> from + (entry-lowram) ).  The result was a number of "Error of type 41"
> messages.  If anyone thinks that the error type number is important, I would
> want to double check it, but I don't really think that it is.
> 
> That's it!
> 
> Rich