Corrections to my prior post, the addressing is no mystery (I was looking at it backwards, assuming the address in the LDT file was the beginning address, but it's actually the address of the end of the line).

Soo.... Row 1 devices start at address 0, Row 2 at 64K (not 32K...), Row 3 at 128K and Row 4 at 196K. That leaves 64K blocks for learned commands.

I've also now seen multiple instaces of button assignments that do not end with the number 3. In fact, it just donned on my what Rene was trying to say, but button assignments range from 1 to probably F. And this makes more sense than my prior button assignment algorithm. The number 3 is actually probably apart of the IRCODE section. So, this would mean a limit of 16 user buttons (excluding LAST PAGE) per page.

I confident that the preprogrammed codes are not reflected at all in the LDT file in any way. I've programmed an entire page of buttons and some use the preprogrammed codes, and others use learned codes. The button assignments with learned codes show up in the LDT file, the others do not. No surprise there.

Besides the annoying "duplicate button" problem (two buttons having the same code even when they are programmed with different codes...) I'm pretty happy with the remote, and besides it gives me something to do lately. :) The only thing I cannot do with the software so far is upload the operating program from the remote, all other uploads/downloads work for me.