Re: Fake cursors.

[Luc Teirlinck]

Kim Storm wrote:

   I don't understand -- unless the cursor is already on the last visible line,
   (recenter -1) will have to scroll the window, right?!

Not in emacs -q.  Yes, if you have customized `scroll-margin' to a
non-zero value.

   So you mean, if the last line currently is only paritally visible,
   (recenter -1) will leave a partially visible line at the bottom of the
   window?

Yes.

   Maybe that's what I've seen (a single invisible pixel line would
   enough to provoke this behaviour).

   > It is the next to last fully _or_ partially visible line.  If there
   > are no partially visible lines (recenter -1) always puts point on the
   > last line.  Do you see anything different from what I described above?

   Probably not, but I think that (recenter -1) should put the cursor on
   the last line, independently of the current contents of the window,
   i.e. it should align the cursor line at the bottom of the window
   (possibly making the first line only partially visible).

   Doesn't that make more sense?

That does not seem completely obvious.  It may look that way in the
case where the behavior you describe is due to a single invisible
pixel line.  But what if there is a single "visible" (sort of,
theoretically) pixel line at the bottom?  In that case, the behavior
you suggest would carry point to what is, for all practical purposes,
a line that is off-screen.  A single C-n will carry one to that line
anyway, with the usual user-customizable scrolling behavior.

Sincerely,

Luc.