bug#15783: 24.3; posn-at-x-y is relative to the buffer area, posn-at-point is relative to the text area.

[Eli Zaretskii]

> From: Alex <agrambot@gmail.com>
> Cc: 椎野 裕樹 <shiino.yuki@gmail.com>,
>   15783@debbugs.gnu.org
> Date: Fri, 17 Nov 2017 01:46:36 -0600
> 
> >> On Emacs 22 and 23, both functions are relative to the buffer area
> >> including the header line.  It's fine.
> >
> > That caused much more grave bugs.
> 
> What kind of bugs did it cause?

See bug#7390, for example.

> It's pretty jarring to have the following return nil when a
> header-line is present:
> 
>   (let ((x-y (posn-x-y (posn-at-point))))
>     (equal x-y (posn-x-y (posn-at-x-y (car x-y)
>                                       (cdr x-y)))))

It's less than ideal, but the alternatives are worse.

> I noticed this since an Emacs package uses posn-x-y and posn-at-x-y to
> try to preserve the point's screen coordinates before and after a
> scroll.  With a header-line, this puts the point one row too high.
> 
> It should probably just let-bind `scroll-preserve-screen-position',
> right?

Probably.

> Still, the above is pretty counterintuitive. If nothing else, I
> think this incompatibility should be highlighted.

Not sure what "highlighted" means in this context.  What are you
suggesting in practice?

> P.S. I noticed that the manual for `posn-at-x-y' states that the WHOLE
> argument affects both X and Y, but the docstring only states that X is
> affected.

Feel free to fix such discrepancies without any discussion, and
thanks.