controlling hollow box cursor

controlling hollow box cursor

[Tak Ota]

I've asked this before and haven't received any comments.  Could
someone comment on the following change?  The purpose of this is to
control the appearance/disappearance of hollow box cursor by the
variable `cursor-in-non-selected-windows'.  You can't comment before
trying thought, because the effect is fairly subjective.

-Tak

*** ../../tmp/emacs-21.2.50/src/xterm.c	Tue Apr  2 07:10:36 2002
--- xterm.c	Tue Apr  2 07:29:28 2002
***************
*** 11677,11683 ****
  						       &new_cursor_width);
  	  if (w->cursor_off_p)
  	    {
! 	      if (new_cursor_type == FILLED_BOX_CURSOR)
  		new_cursor_type = HOLLOW_BOX_CURSOR;
  	      else if (new_cursor_type == BAR_CURSOR && new_cursor_width > 1)
  		new_cursor_width = 1;
--- 11677,11683 ----
  						       &new_cursor_width);
  	  if (w->cursor_off_p)
  	    {
! 	      if (new_cursor_type == FILLED_BOX_CURSOR && cursor_non_selected)
  		new_cursor_type = HOLLOW_BOX_CURSOR;
  	      else if (new_cursor_type == BAR_CURSOR && new_cursor_width > 1)
  		new_cursor_width = 1;

*** ../../tmp/emacs-21.2.50/src/w32term.c	Mon Apr  1 07:47:11 2002
--- w32term.c	Tue Apr  2 07:29:11 2002
***************
*** 9774,9780 ****
  						       &new_cursor_width);
  	  if (w->cursor_off_p)
  	    {
! 	      if (new_cursor_type == FILLED_BOX_CURSOR)
  		new_cursor_type = HOLLOW_BOX_CURSOR;
  	      else if (new_cursor_type == BAR_CURSOR && new_cursor_width > 1)
  		new_cursor_width = 1;
--- 9774,9780 ----
  						       &new_cursor_width);
  	  if (w->cursor_off_p)
  	    {
! 	      if (new_cursor_type == FILLED_BOX_CURSOR && cursor_non_selected)
  		new_cursor_type = HOLLOW_BOX_CURSOR;
  	      else if (new_cursor_type == BAR_CURSOR && new_cursor_width > 1)
  		new_cursor_width = 1;

Re: controlling hollow box cursor

[Kim F. Storm]

Tak Ota <Takaaki.Ota@am.sony.com> writes:

> I've asked this before and haven't received any comments.  Could
> someone comment on the following change?  The purpose of this is to
> control the appearance/disappearance of hollow box cursor by the
> variable `cursor-in-non-selected-windows'.  You can't comment before
> trying thought, because the effect is fairly subjective.

I don't see what those two things have in common, so I don't like
the change!

The generic solution would probably be to add a new variable,
alternate-cursor-type which can be set to e.g. nil or 'hollow to
specify the cursor type to use when blinking the standard cursor-type.

However, this is a little complicated since cursor-type is
buffer-local, so I'd guess alternate-cursor-type would have to be
buffer-local as well.  And we would also have to add the
alternate-cursor-type to the frame-parameters to initialize
the cursor type.

A simpler approach would be to have two global variables:
alternate-block-cursor-type and alternate-bar-cursor-type
which would be used in all buffers.  This could be implemented
in 30 minutes.

However, IIRC, Richard didn't want to make this configurable.

-- 
Kim F. Storm <storm@cua.dk> http://www.cua.dk

Re: controlling hollow box cursor

[Richard Stallman]

    However, IIRC, Richard didn't want to make this configurable.

I don't think it would do any special harm for this to be
configurable, I just have not seen enough reason to think it is worth
the effort.

Re: controlling hollow box cursor

[Tak Ota]

Wed, 3 Apr 2002 15:49:31 -0700 (MST): Richard Stallman <rms@gnu.org> wrote:

>     However, IIRC, Richard didn't want to make this configurable.
> 
> I don't think it would do any special harm for this to be
> configurable, I just have not seen enough reason to think it is worth
> the effort.

The reason depends on the display system you have.  On my system the
hollow box cursor really obscures the character underneath and
currently users must accept it with the blinking cursor.

I proposed that kludge solution, I admit it was, because you mentioned
that making it configurable would do a harm (bloat) last time.

-Tak

Wed, 13 Mar 2002 03:58:34 -0700 (MST): Richard Stallman <rms@gnu.org> wrote:

> Is this user option really necessary?  We should not add a new user
> option just because its definition is meaningful and *someone* might
> want it, because that would make Emacs even more bloated.  We have to
> *try* to keep the growth of user options in check, by adding only
> those that are substantially important--those that we expect
> more than a few users would want.