Re: M-x term off-by-one problem in ESC[xM processing?

[Richard Stallman]

    !     (term-down (- (1+ term-scroll-end) save-current-row lines))

If that really works, can someone figure out why it works?
term-scroll-end is supposed to be the vpos of the first line
that shouldn't move (usually the nonexistent line at end of the screen),
so the old code looks like it should be correct.

So if the code does need this change, it should also have a comment to
explain why add one.