Perhaps we can agree on this:
- PingPong thinks that pages do not deserve this qualification (for its own reasons explained in this thread).
This is not exact:
a) Pages are a concept that V9938 docs somewhere use.
b) This does not mean the V9938 had the concept of pages itself, the v9938 docs have
c) R#2 is not a page oriented register, it's only what it is created for : a register, holding a base memory address
d) as an 
address register it could not be a page register, because it holds an address not a logical page.
e) one could argue that the bit nTh of whatever register could be used to select a page in a specific bitmap mode. And this is true, (not only in bitmap modes) but this abstraction is in the mind of the people that say this. It is not a VDP concept, like for example is the X, Y coordinate used in a logical command register (which in contrast is a TRUE abstraction, a thing that hide hw low level complexity, in this case the complexity of calculating the address of a pixel which is the target of a specific logical vdp command)
Hope this is clear.
It is not that i cannot agree with this definition, it is that even the V9938 define this an an pattern layout address register. This is not by accident.
Except that nobody questions the fact that R#2 bits are defined as pattern table address for the non-bitmap mode.
On the other hand, for all the bitmap modes of the MSX2, the V9938 documentation clearly explains the meaning of the bits as being the page number to be displayed.
You are of course free to disagree with the official V9938 documentation and have your own definition if you wish, but that does not change the facts.
