Re: master b3dd0ce: Provide new option `delete-window-set-selected'

[Eli Zaretskii]

A few comments about this changeset:

Why a variable that is a user option is documented only in the ELisp
manual?  I think it should also be mentioned in the "Change Window"
node of the user manual, where we describe "C-x 0" and "C-x 1".

> +Possible choices are @code{mru} (the default) to select the most
> +recently used window on that frame and @code{pos} to choose the window
                                     ^
There should be a comma before "and" there.

> +at the position of point of the previously selected window.

This part:

  [...] 'pos' to choose the window at the position of point of the
  previously selected window [...]

is confusing: when you say "position of point", the most likely
interpretation is the buffer position of point, which makes the
meaning of the value nonsensical.  I understand that you meant the
frame-relative coordinates of point instead, so the text (here and in
the doc string) should IMO be amended to say that.

> +                                                             If this
> +option is @code{nil}, it means to choose the frame's first window
> +instead.

This text doesn't explain what does "the first window" mean in this
context.

> +           Note that a window with a non-@code{nil}
> +@code{no-other-window} parameter is never chosen.

Doesn't this depend on some argument to the relevant functions?

>  @cindex largest window
> -@defun get-largest-window &optional all-frames dedicated not-selected
> +@defun get-largest-window &optional all-frames dedicated not-selected 
> no-other
>  This function returns the window with the largest area (height times
> -width).  The optional argument @var{all-frames} specifies the windows to
> -search, and has the same meaning as in @code{next-window}.
> -
> -A minibuffer window is never a candidate.  A dedicated window
> -(@pxref{Dedicated Windows}) is never a candidate unless the optional
> -argument @var{dedicated} is non-@code{nil}.  The selected window is not
> -a candidate if the optional argument @var{not-selected} is
> -non-@code{nil}.  If the optional argument @var{not-selected} is
> -non-@code{nil} and the selected window is the only candidate, this
> -function returns @code{nil}.
> -
> -If there are two candidate windows of the same size, this function
> -prefers the one that comes first in the cyclic ordering of windows,
> -starting from the selected window.
> +width).  If there are two candidate windows of the same size, it prefers
> +the one that comes first in the cyclic ordering of windows, starting
> +from the selected window.  The meaning of the arguments is the same as
> +for @code{get-lru-window}.
>  @end defun

This hunk loses the information about minibuffer windows and dedicated
windows, at least.  it also seems to lose the information about when
the selected window isn't a candidate.  Why?

> +(defun window-at-pos (x y &optional frame no-other)
> +  "Return live window at coordinates X, Y on specified FRAME.

A better name for this function is window-at-x-y, IMO.  "Pos" can have
ambiguous interpretations, see above.

> +X and Y are counted in pixels from an origin at 0, 0 of FRAME's
> +native frame.  A coordinate on an edge shared by two windows is
> +attributed to the window on the right (or below).  Return nil if

If you say "FRAME-relative coordinates", doesn't that tell the same
story, just much more succinctly and clearly?

> +(defcustom delete-window-set-selected 'mru

Wouldn't the name delete-window-choose-selected be better?  The doc
string says that much:

> +  "How to choose a frame's selected window after window deletion.

Or maybe delete-selected-window-choose-replacement?

> +               ;; If `delete-window-internal' selected a window with a
> +               ;; non-nil 'no-other-window' parameter as its frame's
> +               ;; selected window, try to choose another one.
> +               (catch 'found
> +                 (walk-window-tree
> +                  (lambda (other)
> +                    (unless (window-parameter other 'no-other-window)
> +                      (set-frame-selected-window frame other)
> +                      (throw 'found t)))
> +                  frame))))

That "try" means that we could sometimes fail, and return a window
with a non-nil 'no-other-window' parameter, right?  If so, the doc
strings, which say such a window will _never_ be returned, are
inaccurate, right?

Thanks.