bug#6609: 23.2; Cursor becomes invisible when put over some faces

bug#6609: 23.2; Cursor becomes invisible when put over some faces

[Deniz Dogan]

Severity: important
Version: 23.2
Package: emacs

I'm using Windows 7 and GNU Emacs 23.2.1 (i386-mingw-nt6.1.7600) of
2010-05-08 on G41R2F1.

Steps to reproduce:
1. Make the following the only contents of your .emacs:
(set-foreground-color "black")
(set-background-color "white")
2. Move the cursor over text with the "default" face.
3. Notice how the cursor is now effectively invisible.

This happens on Windows Vista as well.

It also seems to happen when text has the "button" face, as seen in
e.g. the *Help* buffer sometimes.

-- 
Deniz Dogan

bug#6609: 23.2; Cursor becomes invisible when put over some faces

[Eli Zaretskii]

> From: Deniz Dogan <deniz.a.m.dogan@gmail.com>
> Date: Sun, 11 Jul 2010 13:25:03 +0200
> Cc: 
> 
> I'm using Windows 7 and GNU Emacs 23.2.1 (i386-mingw-nt6.1.7600) of
> 2010-05-08 on G41R2F1.
> 
> Steps to reproduce:
> 1. Make the following the only contents of your .emacs:
> (set-foreground-color "black")
> (set-background-color "white")
> 2. Move the cursor over text with the "default" face.
> 3. Notice how the cursor is now effectively invisible.

What exactly does "effectively invisible" mean?  Do you see it or not?
How about a screenshot?

> This happens on Windows Vista as well.
> 
> It also seems to happen when text has the "button" face, as seen in
> e.g. the *Help* buffer sometimes.

I can reproduce neither of these on Windows XP.  (I don't have access
to W7 or Vista with development environment.)

What does Emacs show if you type "M-x describe-face RET cursor RET"?

Do you have some non-standard color schemes defined on those systems?

bug#6609: 23.2; Cursor becomes invisible when put over some faces

[Deniz Dogan]

2010/7/11 Eli Zaretskii <eliz@gnu.org>:
>> From: Deniz Dogan <deniz.a.m.dogan@gmail.com>
>> Date: Sun, 11 Jul 2010 13:25:03 +0200
>> Cc:
>>
>> I'm using Windows 7 and GNU Emacs 23.2.1 (i386-mingw-nt6.1.7600) of
>> 2010-05-08 on G41R2F1.
>>
>> Steps to reproduce:
>> 1. Make the following the only contents of your .emacs:
>> (set-foreground-color "black")
>> (set-background-color "white")
>> 2. Move the cursor over text with the "default" face.
>> 3. Notice how the cursor is now effectively invisible.
>
> What exactly does "effectively invisible" mean?  Do you see it or not?
> How about a screenshot?
>

I don't see it. I would screenshot it if there was anything to be seen.

>> This happens on Windows Vista as well.
>>
>> It also seems to happen when text has the "button" face, as seen in
>> e.g. the *Help* buffer sometimes.
>
> I can reproduce neither of these on Windows XP.  (I don't have access
> to W7 or Vista with development environment.)
>
> What does Emacs show if you type "M-x describe-face RET cursor RET"?
>

Face: cursor (sample) (customize this face)

Documentation:
Basic face for the cursor color under X.
Note: Other faces cannot inherit from the cursor face.

Defined in `faces.el'.

        Family: unspecified
       Foundry: unspecified
         Width: unspecified
        Height: unspecified
        Weight: unspecified
         Slant: unspecified
    Foreground: unspecified
    Background: unspecified
     Underline: unspecified
      Overline: unspecified
Strike-through: unspecified
           Box: unspecified
       Inverse: unspecified
       Stipple: unspecified
          Font: unspecified
       Fontset: unspecified
       Inherit: unspecified

> Do you have some non-standard color schemes defined on those systems?
>

No, I downloaded Emacs from the official FTP server and just added
set-background-color and set-foreground-color to my Emacs.

I can reproduce it on my system like this:

1. runemacs.exe -Q
2. Evaluate (progn (set-background-color "black")
(set-foreground-color "white"))
3. Notice how the cursor is not displayed when put over e.g. an ending
parenthesis.
4. Notice how the cursor *is* displayed when put over e.g. "progn"
(assuming we're in the *scratch* buffer or any buffer with
lisp-interaction-mode and font locking turned on).

-- 
Deniz Dogan

bug#6609: 23.2; Cursor becomes invisible when put over some faces

[Eli Zaretskii]

> From: Deniz Dogan <deniz.a.m.dogan@gmail.com>
> Date: Sun, 11 Jul 2010 16:13:13 +0200
> Cc: 6609@debbugs.gnu.org
> 
> > Do you have some non-standard color schemes defined on those systems?
> >
> 
> No, I downloaded Emacs from the official FTP server and just added
> set-background-color and set-foreground-color to my Emacs.

I didn't mean in Emacs, I meant Windows color scheme.

bug#6609: 23.2; Cursor becomes invisible when put over some faces

[Eli Zaretskii]

> From: Deniz Dogan <deniz.a.m.dogan@gmail.com>
> Date: Sun, 11 Jul 2010 16:13:13 +0200
> Cc: 6609@debbugs.gnu.org
> 
> > What does Emacs show if you type "M-x describe-face RET cursor RET"?
> >
> 
> Face: cursor (sample) (customize this face)
> 
> Documentation:
> Basic face for the cursor color under X.
> Note: Other faces cannot inherit from the cursor face.
> 
> Defined in `faces.el'.
> 
>         Family: unspecified
>        Foundry: unspecified
>          Width: unspecified
>         Height: unspecified
>         Weight: unspecified
>          Slant: unspecified
>     Foreground: unspecified
>     Background: unspecified

What happens if you evaluate this:

  (set-face-background 'cursor "black")

bug#6609: 23.2; Cursor becomes invisible when put over some faces

[Deniz Dogan]

2010/7/11 Eli Zaretskii <eliz@gnu.org>:
>> From: Deniz Dogan <deniz.a.m.dogan@gmail.com>
>> Date: Sun, 11 Jul 2010 16:13:13 +0200
>> Cc: 6609@debbugs.gnu.org
>>
>> > Do you have some non-standard color schemes defined on those systems?
>> >
>>
>> No, I downloaded Emacs from the official FTP server and just added
>> set-background-color and set-foreground-color to my Emacs.
>
> I didn't mean in Emacs, I meant Windows color scheme.
>

It doesn't seem to matter what Windows themes or color schemes I use.

-- 
Deniz Dogan

bug#6609: 23.2; Cursor becomes invisible when put over some faces

[Deniz Dogan]

2010/7/11 Eli Zaretskii <eliz@gnu.org>:
>> From: Deniz Dogan <deniz.a.m.dogan@gmail.com>
>> Date: Sun, 11 Jul 2010 16:13:13 +0200
>> Cc: 6609@debbugs.gnu.org
>>
>> > What does Emacs show if you type "M-x describe-face RET cursor RET"?
>> >
>>
>> Face: cursor (sample) (customize this face)
>>
>> Documentation:
>> Basic face for the cursor color under X.
>> Note: Other faces cannot inherit from the cursor face.
>>
>> Defined in `faces.el'.
>>
>>         Family: unspecified
>>        Foundry: unspecified
>>          Width: unspecified
>>         Height: unspecified
>>         Weight: unspecified
>>          Slant: unspecified
>>     Foreground: unspecified
>>     Background: unspecified
>
> What happens if you evaluate this:
>
>  (set-face-background 'cursor "black")
>

It returns nil and doesn't seem to have any effect on the problem I'm having.

-- 
Deniz Dogan

bug#6609: 23.2; Cursor becomes invisible when put over some faces

[Deniz Dogan]

2010/7/11 Deniz Dogan <deniz.a.m.dogan@gmail.com>:
> 2010/7/11 Eli Zaretskii <eliz@gnu.org>:
>>> From: Deniz Dogan <deniz.a.m.dogan@gmail.com>
>>> Date: Sun, 11 Jul 2010 16:13:13 +0200
>>> Cc: 6609@debbugs.gnu.org
>>>
>>> > What does Emacs show if you type "M-x describe-face RET cursor RET"?
>>> >
>>>
>>> Face: cursor (sample) (customize this face)
>>>
>>> Documentation:
>>> Basic face for the cursor color under X.
>>> Note: Other faces cannot inherit from the cursor face.
>>>
>>> Defined in `faces.el'.
>>>
>>>         Family: unspecified
>>>        Foundry: unspecified
>>>          Width: unspecified
>>>         Height: unspecified
>>>         Weight: unspecified
>>>          Slant: unspecified
>>>     Foreground: unspecified
>>>     Background: unspecified
>>
>> What happens if you evaluate this:
>>
>>  (set-face-background 'cursor "black")
>>
>
> It returns nil and doesn't seem to have any effect on the problem I'm having.
>

Actually, it seems to make the cursor black, "covering" text with the
default face. So now I can sort of guess where the cursor is by
looking for gaps in the text, but it's of course still not
satisfactory.

-- 
Deniz Dogan

bug#6609: 23.2; Cursor becomes invisible when put over some faces

[Eli Zaretskii]

> From: Deniz Dogan <deniz.a.m.dogan@gmail.com>
> Date: Sun, 11 Jul 2010 19:02:21 +0200
> Cc: 6609@debbugs.gnu.org
> 
> >> What happens if you evaluate this:
> >>
> >>  (set-face-background 'cursor "black")
> >>
> >
> > It returns nil and doesn't seem to have any effect on the problem I'm having.
> >
> 
> Actually, it seems to make the cursor black, "covering" text with the
> default face. So now I can sort of guess where the cursor is by
> looking for gaps in the text, but it's of course still not
> satisfactory.

And if you use a color other than black? does the above has the
expected effect?

IOW, is the problem only with the black color, or with any color, as
far as the `cursor' face is concerned?

bug#6609: 23.2; Cursor becomes invisible when put over some faces

[Deniz Dogan]

2010/7/11 Eli Zaretskii <eliz@gnu.org>:
>> From: Deniz Dogan <deniz.a.m.dogan@gmail.com>
>> Date: Sun, 11 Jul 2010 19:02:21 +0200
>> Cc: 6609@debbugs.gnu.org
>>
>> >> What happens if you evaluate this:
>> >>
>> >>  (set-face-background 'cursor "black")
>> >>
>> >
>> > It returns nil and doesn't seem to have any effect on the problem I'm having.
>> >
>>
>> Actually, it seems to make the cursor black, "covering" text with the
>> default face. So now I can sort of guess where the cursor is by
>> looking for gaps in the text, but it's of course still not
>> satisfactory.
>
> And if you use a color other than black? does the above has the
> expected effect?
>
> IOW, is the problem only with the black color, or with any color, as
> far as the `cursor' face is concerned?
>

Setting it to white or red or yellow gives me the desired effect. Thanks.

But this still should not be needed, so I'd like it if you don't close this bug.

-- 
Deniz Dogan

bug#6609: 23.2; Cursor becomes invisible when put over some faces

[Eli Zaretskii]

> From: Deniz Dogan <deniz.a.m.dogan@gmail.com>
> Date: Sun, 11 Jul 2010 19:24:48 +0200
> Cc: 6609@debbugs.gnu.org
> 
> 2010/7/11 Eli Zaretskii <eliz@gnu.org>:
> >> From: Deniz Dogan <deniz.a.m.dogan@gmail.com>
> >> Date: Sun, 11 Jul 2010 19:02:21 +0200
> >> Cc: 6609@debbugs.gnu.org
> >>
> >> >> What happens if you evaluate this:
> >> >>
> >> >>  (set-face-background 'cursor "black")
> >> >>
> >> >
> >> > It returns nil and doesn't seem to have any effect on the problem I'm having.
> >> >
> >>
> >> Actually, it seems to make the cursor black, "covering" text with the
> >> default face. So now I can sort of guess where the cursor is by
> >> looking for gaps in the text, but it's of course still not
> >> satisfactory.
> >
> > And if you use a color other than black? does the above has the
> > expected effect?
> >
> > IOW, is the problem only with the black color, or with any color, as
> > far as the `cursor' face is concerned?
> >
> 
> Setting it to white or red or yellow gives me the desired effect. Thanks.
> 
> But this still should not be needed, so I'd like it if you don't close this bug.

Who said anything about closing?  I'm just warming up.

So let's back up a bit and see if we have the whole picture:

 . What happens in "emacs -Q" without evaluating

     (set-foreground-color "black")
     (set-background-color "white")

   ?  Is the cursor visible?  Is it black?

 . Which one of the above expressions triggers the problem when
   evaluated?  Is the order of evaluating them important?

 . When you set the background color of the `cursor' face, does the
   problem happen only with the black color?  What about colors close
   to black, like "gray1", "gray2", etc.? how about "gray40" or
   "gray80"?

bug#6609: 23.2; Cursor becomes invisible when put over some faces

[Eli Zaretskii]

> From: Deniz Dogan <deniz.a.m.dogan@gmail.com>
> Date: Sun, 11 Jul 2010 16:38:02 +0200
> Cc: 6609@debbugs.gnu.org
> 
> > I didn't mean in Emacs, I meant Windows color scheme.
> >
> 
> It doesn't seem to matter what Windows themes or color schemes I use.

What about the screen resolution and/or the number of colors? do any
of them have any effect?

bug#6609: 23.2; Cursor becomes invisible when put over some faces

[Deniz Dogan]

[-- Attachment #1: Type: text/plain, Size: 1789 bytes --]

2010/7/11 Eli Zaretskii <eliz@gnu.org>:
> So let's back up a bit and see if we have the whole picture:
>
>  . What happens in "emacs -Q" without evaluating
>
>     (set-foreground-color "black")
>     (set-background-color "white")
>
>   ?  Is the cursor visible?  Is it black?
>

It is blinking, visible and black.

>  . Which one of the above expressions triggers the problem when
>   evaluated?  Is the order of evaluating them important?
>

Assuming you meant (set-foreground-color "white") and
(set-background-color "black")...

Scenario 1:
1. I set the foreground to white, everything remains fine.
2. I set the background to black, the problem as originally described
happens, i.e. the cursor is invisible on default face text. I cannot
see anything blinking.

Scenario 2:
1. I set the background to black. The cursor *looks* invisible on
default text every other blink, but this is expected, since the text
is black on black background.
2. I set the foreground to white, the problem as originally described happens.

>  . When you set the background color of the `cursor' face, does the
>   problem happen only with the black color?  What about colors close
>   to black, like "gray1", "gray2", etc.? how about "gray40" or
>   "gray80"?
>

Assuming white foreground and black background...

I set the cursor color to #333333, which is nearly black but still
visible on black background. During the time period at which the
cursor is not visible (the first half of the blinking), everything
looks fine. When the cursor is visible (the other half of the
blinking), the text which is normally white (default face) appears
black. I can see the #333333 cursor behind the text. I have attached a
GIF screenshot of this.

-- 
Deniz Dogan

[-- Attachment #2: color333.gif --]
[-- Type: image/gif, Size: 1916 bytes --]

bug#6609: 23.2; Cursor becomes invisible when put over some faces

[Deniz Dogan]

2010/7/11 Eli Zaretskii <eliz@gnu.org>:
>> From: Deniz Dogan <deniz.a.m.dogan@gmail.com>
>> Date: Sun, 11 Jul 2010 16:38:02 +0200
>> Cc: 6609@debbugs.gnu.org
>>
>> > I didn't mean in Emacs, I meant Windows color scheme.
>> >
>>
>> It doesn't seem to matter what Windows themes or color schemes I use.
>
> What about the screen resolution and/or the number of colors? do any
> of them have any effect?
>

I tried lowering both of them to 16 bit and 1280x760 but neither of
these tricks seemed to help the problem.

I normally use 32 bit and 1920x1080.

-- 
Deniz Dogan

bug#6609: 23.2; Cursor becomes invisible when put over some faces

[Eli Zaretskii]

> From: Deniz Dogan <deniz.a.m.dogan@gmail.com>
> Date: Sun, 11 Jul 2010 20:24:04 +0200
> Cc: 6609@debbugs.gnu.org
> 
> 2010/7/11 Eli Zaretskii <eliz@gnu.org>:
> > So let's back up a bit and see if we have the whole picture:
> >
> >  . What happens in "emacs -Q" without evaluating
> >
> >     (set-foreground-color "black")
> >     (set-background-color "white")
> >
> >   ?  Is the cursor visible?  Is it black?
> >
> 
> It is blinking, visible and black.
> 
> >  . Which one of the above expressions triggers the problem when
> >   evaluated?  Is the order of evaluating them important?
> >
> 
> Assuming you meant (set-foreground-color "white") and
> (set-background-color "black")...

No, I meant what you wrote in your original report: black foreground
on white background.  This happens to be the default colors, btw.

Since you evidently now tried the colors in reverse, is it possible
that your original report mistakenly swapped the colors, and what you
_really_ meant was white foreground on black background?  Could you
please go back and validate your original report?

bug#6609: 23.2; Cursor becomes invisible when put over some faces

[Deniz Dogan]

2010/7/11 Eli Zaretskii <eliz@gnu.org>:
>> From: Deniz Dogan <deniz.a.m.dogan@gmail.com>
>> Date: Sun, 11 Jul 2010 20:24:04 +0200
>> Cc: 6609@debbugs.gnu.org
>>
>> 2010/7/11 Eli Zaretskii <eliz@gnu.org>:
>> > So let's back up a bit and see if we have the whole picture:
>> >
>> >  . What happens in "emacs -Q" without evaluating
>> >
>> >     (set-foreground-color "black")
>> >     (set-background-color "white")
>> >
>> >   ?  Is the cursor visible?  Is it black?
>> >
>>
>> It is blinking, visible and black.
>>
>> >  . Which one of the above expressions triggers the problem when
>> >   evaluated?  Is the order of evaluating them important?
>> >
>>
>> Assuming you meant (set-foreground-color "white") and
>> (set-background-color "black")...
>
> No, I meant what you wrote in your original report: black foreground
> on white background.  This happens to be the default colors, btw.
>
> Since you evidently now tried the colors in reverse, is it possible
> that your original report mistakenly swapped the colors, and what you
> _really_ meant was white foreground on black background?  Could you
> please go back and validate your original report?
>

Terribly sorry about that. I did mean white foreground on black
background in the original report!

bug#6609: 23.2; Cursor becomes invisible when put over some faces

[Eli Zaretskii]

> From: Deniz Dogan <deniz.a.m.dogan@gmail.com>
> Date: Sun, 11 Jul 2010 21:49:12 +0200
> Cc: 6609@debbugs.gnu.org
> 
> Terribly sorry about that. I did mean white foreground on black
> background in the original report!

A world of difference.  I do see this problem on XP as well, then.

Can someone try this on GNU/Linux in a GUI session, please?  I'd like
to know if this problem is specific to w32.

To repeat the recipe:

  emacs -Q

Insert this in the *scratch* buffer:

  (set-background-color "black")
  (set-foreground-color "white")

Evaluate these two lines.  Your cursor will now become invisible.  You
can only see it (a) when it is on some text of a non-default color, or
(b) when the Emacs frame does not have focus, and the cursor is
displayed as a hollow rectangle.

"M-x set-face-background RET cursor RET white RET" fixes the problem.
So my first guess would be that we don't do that automatically as part
of changing the default face, for some reason.

bug#6609: 23.2; Cursor becomes invisible when put over some faces

[Deniz Dogan]

2010/7/12 Eli Zaretskii <eliz@gnu.org>:
>> From: Deniz Dogan <deniz.a.m.dogan@gmail.com>
>> Date: Sun, 11 Jul 2010 21:49:12 +0200
>> Cc: 6609@debbugs.gnu.org
>>
>> Terribly sorry about that. I did mean white foreground on black
>> background in the original report!
>
> A world of difference.  I do see this problem on XP as well, then.
>
> Can someone try this on GNU/Linux in a GUI session, please?  I'd like
> to know if this problem is specific to w32.
>
> To repeat the recipe:
>
>  emacs -Q
>
> Insert this in the *scratch* buffer:
>
>  (set-background-color "black")
>  (set-foreground-color "white")
>
> Evaluate these two lines.  Your cursor will now become invisible.  You
> can only see it (a) when it is on some text of a non-default color, or
> (b) when the Emacs frame does not have focus, and the cursor is
> displayed as a hollow rectangle.
>
> "M-x set-face-background RET cursor RET white RET" fixes the problem.
> So my first guess would be that we don't do that automatically as part
> of changing the default face, for some reason.
>

I have tried this on GNU/Linux as well (Bazaar version from some time
in June) and the problem does not appear there. I can't remember it
ever happening on GNU/Linux actually.

bug#6609: 23.2; Cursor becomes invisible when put over some faces

[Eli Zaretskii]

> From: Deniz Dogan <deniz.a.m.dogan@gmail.com>
> Date: Mon, 12 Jul 2010 10:09:04 +0200
> Cc: 6609@debbugs.gnu.org
> 
> 2010/7/12 Eli Zaretskii <eliz@gnu.org>:
> >> From: Deniz Dogan <deniz.a.m.dogan@gmail.com>
> >> Date: Sun, 11 Jul 2010 21:49:12 +0200
> >> Cc: 6609@debbugs.gnu.org
> >>
> >> Terribly sorry about that. I did mean white foreground on black
> >> background in the original report!
> >
> > A world of difference.  I do see this problem on XP as well, then.
> >
> > Can someone try this on GNU/Linux in a GUI session, please?  I'd like
> > to know if this problem is specific to w32.
> >
> > To repeat the recipe:
> >
> >  emacs -Q
> >
> > Insert this in the *scratch* buffer:
> >
> >  (set-background-color "black")
> >  (set-foreground-color "white")
> >
> > Evaluate these two lines.  Your cursor will now become invisible.  You
> > can only see it (a) when it is on some text of a non-default color, or
> > (b) when the Emacs frame does not have focus, and the cursor is
> > displayed as a hollow rectangle.
> >
> > "M-x set-face-background RET cursor RET white RET" fixes the problem.
> > So my first guess would be that we don't do that automatically as part
> > of changing the default face, for some reason.
> >
> 
> I have tried this on GNU/Linux as well (Bazaar version from some time
> in June) and the problem does not appear there. I can't remember it
> ever happening on GNU/Linux actually.

Can you please try the patch below?  If it works for you, I will
commit it to the emacs-23 branch.

=== modified file 'src/w32fns.c'
--- src/w32fns.c	2010-07-10 18:52:53 +0000
+++ src/w32fns.c	2010-07-12 10:31:49 +0000
@@ -1333,7 +1333,10 @@ x_set_foreground_color (struct frame *f,
   if (FRAME_W32_WINDOW (f) != 0)
     {
       if (x->cursor_pixel == old_fg)
-	x->cursor_pixel = fg;
+	{
+	  x->cursor_pixel = fg;
+	  x->cursor_gc->background = fg;
+	}
 
       update_face_from_frame_parameter (f, Qforeground_color, arg);
       if (FRAME_VISIBLE_P (f))

bug#6609: 23.2; Cursor becomes invisible when put over some faces

[Deniz Dogan]

2010/7/12 Eli Zaretskii <eliz@gnu.org>:
>> From: Deniz Dogan <deniz.a.m.dogan@gmail.com>
>> Date: Mon, 12 Jul 2010 10:09:04 +0200
>> Cc: 6609@debbugs.gnu.org
>>
>> 2010/7/12 Eli Zaretskii <eliz@gnu.org>:
>> >> From: Deniz Dogan <deniz.a.m.dogan@gmail.com>
>> >> Date: Sun, 11 Jul 2010 21:49:12 +0200
>> >> Cc: 6609@debbugs.gnu.org
>> >>
>> >> Terribly sorry about that. I did mean white foreground on black
>> >> background in the original report!
>> >
>> > A world of difference.  I do see this problem on XP as well, then.
>> >
>> > Can someone try this on GNU/Linux in a GUI session, please?  I'd like
>> > to know if this problem is specific to w32.
>> >
>> > To repeat the recipe:
>> >
>> >  emacs -Q
>> >
>> > Insert this in the *scratch* buffer:
>> >
>> >  (set-background-color "black")
>> >  (set-foreground-color "white")
>> >
>> > Evaluate these two lines.  Your cursor will now become invisible.  You
>> > can only see it (a) when it is on some text of a non-default color, or
>> > (b) when the Emacs frame does not have focus, and the cursor is
>> > displayed as a hollow rectangle.
>> >
>> > "M-x set-face-background RET cursor RET white RET" fixes the problem.
>> > So my first guess would be that we don't do that automatically as part
>> > of changing the default face, for some reason.
>> >
>>
>> I have tried this on GNU/Linux as well (Bazaar version from some time
>> in June) and the problem does not appear there. I can't remember it
>> ever happening on GNU/Linux actually.
>
> Can you please try the patch below?  If it works for you, I will
> commit it to the emacs-23 branch.
>
> === modified file 'src/w32fns.c'
> --- src/w32fns.c        2010-07-10 18:52:53 +0000
> +++ src/w32fns.c        2010-07-12 10:31:49 +0000
> @@ -1333,7 +1333,10 @@ x_set_foreground_color (struct frame *f,
>   if (FRAME_W32_WINDOW (f) != 0)
>     {
>       if (x->cursor_pixel == old_fg)
> -       x->cursor_pixel = fg;
> +       {
> +         x->cursor_pixel = fg;
> +         x->cursor_gc->background = fg;
> +       }
>
>       update_face_from_frame_parameter (f, Qforeground_color, arg);
>       if (FRAME_VISIBLE_P (f))
>
>

I don't build my Emacs on Windows, only on GNU/Linux. Could someone
else running Windows try it? Thanks.

bug#6609: 23.2; Cursor becomes invisible when put over some faces

[Eli Zaretskii]

> From: Deniz Dogan <deniz.a.m.dogan@gmail.com>
> Date: Mon, 12 Jul 2010 12:47:15 +0200
> Cc: 6609@debbugs.gnu.org
> 
> > --- src/w32fns.c        2010-07-10 18:52:53 +0000
> > +++ src/w32fns.c        2010-07-12 10:31:49 +0000
> > @@ -1333,7 +1333,10 @@ x_set_foreground_color (struct frame *f,
> >   if (FRAME_W32_WINDOW (f) != 0)
> >     {
> >       if (x->cursor_pixel == old_fg)
> > -       x->cursor_pixel = fg;
> > +       {
> > +         x->cursor_pixel = fg;
> > +         x->cursor_gc->background = fg;
> > +       }
> >
> >       update_face_from_frame_parameter (f, Qforeground_color, arg);
> >       if (FRAME_VISIBLE_P (f))
> >
> >
> 
> I don't build my Emacs on Windows, only on GNU/Linux. Could someone
> else running Windows try it? Thanks.

I already tried this on my Windows XP machine, and it does fix the
problem.  But I wanted to be sure it works on Vista and W7 as well.

bug#6609: 23.2; Cursor becomes invisible when put over some faces

[Juanma Barranquero]

On Mon, Jul 12, 2010 at 14:28, Eli Zaretskii <eliz@gnu.org> wrote:

> I already tried this on my Windows XP machine, and it does fix the
> problem.  But I wanted to be sure it works on Vista and W7 as well.

Yes, with your patch the cursor is visible on Windows 7 Home Premium.

    Juanma

bug#6609: 23.2; Cursor becomes invisible when put over some faces

[Eli Zaretskii]

> From: Juanma Barranquero <lekktu@gmail.com>
> Date: Wed, 14 Jul 2010 01:43:47 +0200
> Cc: Deniz Dogan <deniz.a.m.dogan@gmail.com>, 6609@debbugs.gnu.org
> 
> On Mon, Jul 12, 2010 at 14:28, Eli Zaretskii <eliz@gnu.org> wrote:
> 
> > I already tried this on my Windows XP machine, and it does fix the
> > problem.  But I wanted to be sure it works on Vista and W7 as well.
> 
> Yes, with your patch the cursor is visible on Windows 7 Home Premium.

Thanks, now installed on the emacs-23 branch (revno 99924).