23.0.50; display property and continuation arrows

23.0.50; display property and continuation arrows

[Stephen Berman]

GNU Emacs 23.0.50.1 (i686-pc-linux-gnu, GTK+ Version 2.12.0) of
2007-12-09 on escher

1. emacs -Q

2. Type a line of text that extends beyond (window-width) with auto-fill
mode is disabled so that continuation arrows are displayed in the
fringes.

3. Put the cursor on the second column after the left continuation arrow
and evaluate this sexp: (put-text-property (1- (point)) (point) 'display
" \n").

4. Now the rest of the text, from point to the end, appears to have
moved down by one line, i.e., the continuation arrows remain but after
the left continuation arrow it looks like there is an empty line before
the rest of the text.

Compare this with the following step instead of 3:

3'. Put the cursor on the first column after the left continuation arrow
and evaluate this sexp: (put-text-property (1- (point)) (point) 'display
" \n").

5. Now the rest of the text, from point to the end, appears in the same
position as before step 3' but the continuation arrows have vanished,
i.e., it looks like there are two lines of text, not one continued line.

Is the difference between 4 and 5, resulting from the different
positions of the display property, expected?  If so, is it conceivable
to change the display code so that doing step 3 (not step 3') would
result in the appearance in 5 instead of that in 4?

Steve Berman

Re: 23.0.50; display property and continuation arrows

[Richard Stallman]

    3'. Put the cursor on the first column after the left continuation arrow
    and evaluate this sexp: (put-text-property (1- (point)) (point) 'display
    " \n").

    5. Now the rest of the text, from point to the end, appears in the same
    position as before step 3' but the continuation arrows have vanished,
    i.e., it looks like there are two lines of text, not one continued line.

When I try this, the continuation arrows do not disappear.

I am not sure that there is a bug here.

Re: 23.0.50; display property and continuation arrows

[Stephen Berman]

[-- Attachment #1: Type: text/plain, Size: 884 bytes --]

On Sun, 23 Dec 2007 16:12:01 -0500 Richard Stallman <rms@gnu.org> wrote:

>     3'. Put the cursor on the first column after the left continuation arrow
>     and evaluate this sexp: (put-text-property (1- (point)) (point) 'display
>     " \n").
>
>     5. Now the rest of the text, from point to the end, appears in the same
>     position as before step 3' but the continuation arrows have vanished,
>     i.e., it looks like there are two lines of text, not one continued line.
>
> When I try this, the continuation arrows do not disappear.

This surprises me.  I continue to see what I reported with current
sources, GNU Emacs 23.0.50.2 (i686-pc-linux-gnu, GTK+ Version 2.12.0) of
2008-01-03 on escher.  Perhaps the steps I gave to reproduce it were not
clear.  I've made screen shots showing what I see.  Here is the buffer
at step 3 of my OP, before evalling the display sexp:


[-- Attachment #2: continuation1 --]
[-- Type: image/png, Size: 3477 bytes --]

[-- Attachment #3: Type: text/plain, Size: 127 bytes --]


Note the cursor is over the "e" of the last occurrence of "testing".
And here is the buffer after evalling the display sexp:


[-- Attachment #4: continuation2 --]
[-- Type: image/png, Size: 4847 bytes --]

[-- Attachment #5: Type: text/plain, Size: 150 bytes --]


Note the continuation arrows and the appearance of an "empty" line.
Next, here is the buffer at step 3' of my OP, before evalling the
display sexp:


[-- Attachment #6: continuation3 --]
[-- Type: image/png, Size: 4756 bytes --]

[-- Attachment #7: Type: text/plain, Size: 164 bytes --]


Note the cursor is over the first "t" of the last occurrence of
"testing", immediately after the fringe.  And here is the buffer after
evalling the display sexp:


[-- Attachment #8: continuation4 --]
[-- Type: image/png, Size: 5981 bytes --]

[-- Attachment #9: Type: text/plain, Size: 712 bytes --]


Note the absence of the continuation arrows and of the "empty" line.
Are you unable to reproduce this?

> I am not sure that there is a bug here.

I don't know if it is a bug, because the results seen above may be
consistent with the code (I cannot tell).  My OP was a request for a
explanation of the difference seen in the second and fourth images, as
well as (if it is not a bug) a feature request: I would like *not* to
have continuation arrows and an "empty" line in the second image.  In
other words, I would like the appearance in the second image to be like
that in the last image (modulo the cursor position and the position of
the display property).  Is this possible?  If not, why not?

Steve Berman

[-- Attachment #10: Type: text/plain, Size: 142 bytes --]

_______________________________________________
Emacs-devel mailing list
Emacs-devel@gnu.org
http://lists.gnu.org/mailman/listinfo/emacs-devel

Re: 23.0.50; display property and continuation arrows

[martin rudalics]

>     This surprises me.  I continue to see what I reported with current
>     sources, GNU Emacs 23.0.50.2 (i686-pc-linux-gnu, GTK+ Version 2.12.0) of
>     2008-01-03 on escher.  Perhaps the steps I gave to reproduce it were not
>     clear.  I've made screen shots showing what I see.  Here is the buffer
>     at step 3 of my OP, before evalling the display sexp:
> 
> Maybe it only fails with the GTK build.  Jan, can you reproduce the problem?

Reproducible here on Windows (without GTK).

Re: 23.0.50; display property and continuation arrows

[Jan Djärv]

Richard Stallman skrev:
>     This surprises me.  I continue to see what I reported with current
>     sources, GNU Emacs 23.0.50.2 (i686-pc-linux-gnu, GTK+ Version 2.12.0) of
>     2008-01-03 on escher.  Perhaps the steps I gave to reproduce it were not
>     clear.  I've made screen shots showing what I see.  Here is the buffer
>     at step 3 of my OP, before evalling the display sexp:
> 
> Maybe it only fails with the GTK build.  Jan, can you reproduce the problem?
> 

Yes I cam.  On different toolkit builds, so it looks to be tool kit independent.

	Jan D.

Re: 23.0.50; display property and continuation arrows

[Jan Djärv]

Jan Djärv skrev:
> Richard Stallman skrev:
>>     This surprises me.  I continue to see what I reported with current
>>     sources, GNU Emacs 23.0.50.2 (i686-pc-linux-gnu, GTK+ Version 2.12.0) of
>>     2008-01-03 on escher.  Perhaps the steps I gave to reproduce it were not
>>     clear.  I've made screen shots showing what I see.  Here is the buffer
>>     at step 3 of my OP, before evalling the display sexp:
>>
>> Maybe it only fails with the GTK build.  Jan, can you reproduce the problem?
>>
> 
> Yes I cam.  On different toolkit builds, so it looks to be tool kit independent.
> 

A futher observation.  If the character in the second column is a space, the
bug does not appear.

	Jan D.

Re: 23.0.50; display property and continuation arrows

[Stephen Berman]

On Sat, 05 Jan 2008 18:38:59 +0100 Jan Djärv <jan.h.d@swipnet.se> wrote:

> Richard Stallman skrev:
>>     This surprises me.  I continue to see what I reported with current
>>     sources, GNU Emacs 23.0.50.2 (i686-pc-linux-gnu, GTK+ Version 2.12.0) of
>>     2008-01-03 on escher.  Perhaps the steps I gave to reproduce it were not
>>     clear.  I've made screen shots showing what I see.  Here is the buffer
>>     at step 3 of my OP, before evalling the display sexp:
>> 
>> Maybe it only fails with the GTK build.  Jan, can you reproduce the problem?
>> 
>
> Yes I cam.  On different toolkit builds, so it looks to be tool kit independent.

It is definitely toolkit independent, because a similar effect obtains
with emacs -Q -nw: when the display property is added with the cursor on
the second column of the continued line, the continuation slash at the
end of the "first" line remains and there appears to be an empty line
(cf. the second image of my previous post); but the display property is
added with the cursor on the first column of the continued line, the
continuation slash at the end of the "first" line vanishes and there is
no "empty" line (cf. the fourth image of my previous post).

Steve Berman

Re: 23.0.50; display property and continuation arrows

[Stephen Berman]

On Sat, 05 Jan 2008 18:40:43 +0100 Jan Djärv <jan.h.d@swipnet.se> wrote:

> Jan Djärv skrev:
>> Richard Stallman skrev:
>>>     This surprises me.  I continue to see what I reported with current
>>>     sources, GNU Emacs 23.0.50.2 (i686-pc-linux-gnu, GTK+ Version 2.12.0) of
>>>     2008-01-03 on escher.  Perhaps the steps I gave to reproduce it were not
>>>     clear.  I've made screen shots showing what I see.  Here is the buffer
>>>     at step 3 of my OP, before evalling the display sexp:
>>>
>>> Maybe it only fails with the GTK build.  Jan, can you reproduce the problem?
>>>
>> 
>> Yes I cam.  On different toolkit builds, so it looks to be tool kit independent.
>> 
>
> A futher observation.  If the character in the second column is a space, the
> bug does not appear.

I'm not sure what you mean: do you mean the position occupied by "e" in
the last occurrence of "testing" in my screen first shot?  If I make
this a space, I see no different behavior.  Also, what do you mean by
the "bug"?

Steve Berman

Re: 23.0.50; display property and continuation arrows

[Jan Djärv]

Stephen Berman skrev:
> On Sat, 05 Jan 2008 18:40:43 +0100 Jan Djärv <jan.h.d@swipnet.se> wrote:
> 
>> Jan Djärv skrev:
>>> Richard Stallman skrev:
>>>>     This surprises me.  I continue to see what I reported with current
>>>>     sources, GNU Emacs 23.0.50.2 (i686-pc-linux-gnu, GTK+ Version 2.12.0) of
>>>>     2008-01-03 on escher.  Perhaps the steps I gave to reproduce it were not
>>>>     clear.  I've made screen shots showing what I see.  Here is the buffer
>>>>     at step 3 of my OP, before evalling the display sexp:
>>>>
>>>> Maybe it only fails with the GTK build.  Jan, can you reproduce the problem?
>>>>
>>> Yes I cam.  On different toolkit builds, so it looks to be tool kit independent.
>>>
>> A futher observation.  If the character in the second column is a space, the
>> bug does not appear.
> 
> I'm not sure what you mean: do you mean the position occupied by "e" in
> the last occurrence of "testing" in my screen first shot?  If I make
> this a space, I see no different behavior.  Also, what do you mean by
> the "bug"?

The bug == the appearance of an "empty" line.  Yes, the e in your screen shot.

	Jan D.

Re: 23.0.50; display property and continuation arrows

[Stephen Berman]

[-- Attachment #1: Type: text/plain, Size: 1068 bytes --]

On Sat, 05 Jan 2008 23:59:07 +0100 Jan Djärv <jan.h.d@swipnet.se> wrote:

> Stephen Berman skrev:
>> On Sat, 05 Jan 2008 18:40:43 +0100 Jan Djärv <jan.h.d@swipnet.se> wrote:
[...]
>>> A futher observation.  If the character in the second column is a space, the
>>> bug does not appear.
>>
>> I'm not sure what you mean: do you mean the position occupied by "e" in
>> the last occurrence of "testing" in my screen first shot?  If I make
>> this a space, I see no different behavior.  Also, what do you mean by
>> the "bug"?
>
> The bug == the appearance of an "empty" line.  Yes, the e in your screen shot.

Ok, but in this case there is no "empty" line only if you apply the
display property with the cursor over the "s" following the space,
i.e. the cursor is on the third column.  But that's also so if you leave
the "e", so the presence of the space is irrelevant.

What I'm really interested in is cases where the first column of the
continued line is a space and the last column of the first part of the
line contains a nonspace character, like the following:


[-- Attachment #2: continuation5 --]
[-- Type: image/png, Size: 6694 bytes --]

[-- Attachment #3: Type: text/plain, Size: 90 bytes --]


Now, if the display property is applied to the second column, the result
is as follows:


[-- Attachment #4: continuation6 --]
[-- Type: image/png, Size: 6866 bytes --]

[-- Attachment #5: Type: text/plain, Size: 50 bytes --]


But what I'd like to see as the result is this:


[-- Attachment #6: continuation7 --]
[-- Type: image/png, Size: 7792 bytes --]

[-- Attachment #7: Type: text/plain, Size: 196 bytes --]


which is similar to what results if the display property is applied to
the first column, but that also deletes the last character of the first
part of the line, which I don't want.

Steve Berman

[-- Attachment #8: Type: text/plain, Size: 142 bytes --]

_______________________________________________
Emacs-devel mailing list
Emacs-devel@gnu.org
http://lists.gnu.org/mailman/listinfo/emacs-devel

Re: 23.0.50; display property and continuation arrows

[Jan Djärv]

Stephen Berman skrev:
> On Sat, 05 Jan 2008 23:59:07 +0100 Jan Djärv <jan.h.d@swipnet.se> wrote:
> 
>> Stephen Berman skrev:
>>> On Sat, 05 Jan 2008 18:40:43 +0100 Jan Djärv <jan.h.d@swipnet.se> wrote:
> [...]
>>>> A futher observation.  If the character in the second column is a space, the
>>>> bug does not appear.
>>> I'm not sure what you mean: do you mean the position occupied by "e" in
>>> the last occurrence of "testing" in my screen first shot?  If I make
>>> this a space, I see no different behavior.  Also, what do you mean by
>>> the "bug"?
>> The bug == the appearance of an "empty" line.  Yes, the e in your screen shot.
> 
> Ok, but in this case there is no "empty" line only if you apply the
> display property with the cursor over the "s" following the space,
> i.e. the cursor is on the third column.  But that's also so if you leave
> the "e", so the presence of the space is irrelevant.

That is not what I did.  I had the cursor over the space, in the second 
column.  Maybe word breaks and/or column width has something to do with this also.

	Jan D.

Re: 23.0.50; display property and continuation arrows

[Stephen Berman]

On Sun, 06 Jan 2008 11:46:30 +0100 Jan Djärv <jan.h.d@swipnet.se> wrote:

> Stephen Berman skrev:
>> On Sat, 05 Jan 2008 23:59:07 +0100 Jan Djärv <jan.h.d@swipnet.se> wrote:
>>
>>> Stephen Berman skrev:
>>>> On Sat, 05 Jan 2008 18:40:43 +0100 Jan Djärv <jan.h.d@swipnet.se> wrote:
>> [...]
>>>>> A futher observation.  If the character in the second column is a space, the
>>>>> bug does not appear.
>>>> I'm not sure what you mean: do you mean the position occupied by "e" in
>>>> the last occurrence of "testing" in my screen first shot?  If I make
>>>> this a space, I see no different behavior.  Also, what do you mean by
>>>> the "bug"?
>>> The bug == the appearance of an "empty" line.  Yes, the e in your screen shot.
>>
>> Ok, but in this case there is no "empty" line only if you apply the
>> display property with the cursor over the "s" following the space,
>> i.e. the cursor is on the third column.  But that's also so if you leave
>> the "e", so the presence of the space is irrelevant.
>
> That is not what I did.  I had the cursor over the space, in the
> second column.  Maybe word breaks and/or column width has something to
> do with this also.

Did you use the same recipe I gave in my OP, aside from the space?  With
that and the space I still get the "empty" line, so as you say there
must be some other difference.  When I do emacs -Q and eval
(frame-parameters), I get this:

((parent-id . 25170646)
 (explicit-name)
 (display . ":0")
 (visibility . t)
 (icon-name)
 (outer-window-id . "65011885")
 (window-id . "65011940")
 (top . 30)
 (left . 0)
 (buried-buffer-list)
 (buffer-list #<buffer *scratch*> #<buffer  *Minibuf-1*> #<buffer *Help*>)
 (unsplittable)
 (minibuffer . #<window 4 on  *Minibuf-0*>)
 (modeline . t)
 (width . 80)
 (height . 40)
 (name . "emacs@escher.local.home")
 (environment)
 (background-mode . light)
 (display-type . color)
 (horizontal-scroll-bars . t)
 (window-system . x)
 (scroll-bar-width . 16)
 (cursor-type . box)
 (auto-lower)
 (auto-raise)
 (icon-type . t)
 (wait-for-wm . t)
 (title)
 (buffer-predicate)
 (tool-bar-lines . 1)
 (menu-bar-lines . 1)
 (scroll-bar-background . "grey75")
 (scroll-bar-foreground . "#000000")
 (right-fringe . 8)
 (left-fringe . 8)
 (line-spacing)
 (screen-gamma)
 (border-color . "black")
 (cursor-color . "#000000")
 (mouse-color . "#000000")
 (background-color . "#ffffff")
 (foreground-color . "#000000")
 (vertical-scroll-bars . left)
 (internal-border-width . 0)
 (border-width . 0)
 (font . "-*-fixed-medium-r-*--16-*-*-*-*-*-fontset-16"))

Do you have different values for relevant parameters (e.g. font)?

Steve Berman

Re: 23.0.50; display property and continuation arrows

[Jan Djärv]

Stephen Berman skrev:
> On Sun, 06 Jan 2008 11:46:30 +0100 Jan Djärv <jan.h.d@swipnet.se> wrote:
> 
>> Stephen Berman skrev:
>>> On Sat, 05 Jan 2008 23:59:07 +0100 Jan Djärv <jan.h.d@swipnet.se> wrote:
>>>
>>>> Stephen Berman skrev:
>>>>> On Sat, 05 Jan 2008 18:40:43 +0100 Jan Djärv <jan.h.d@swipnet.se> wrote:
>>> [...]
>>>>>> A futher observation.  If the character in the second column is a space, the
>>>>>> bug does not appear.
>>>>> I'm not sure what you mean: do you mean the position occupied by "e" in
>>>>> the last occurrence of "testing" in my screen first shot?  If I make
>>>>> this a space, I see no different behavior.  Also, what do you mean by
>>>>> the "bug"?
>>>> The bug == the appearance of an "empty" line.  Yes, the e in your screen shot.
>>> Ok, but in this case there is no "empty" line only if you apply the
>>> display property with the cursor over the "s" following the space,
>>> i.e. the cursor is on the third column.  But that's also so if you leave
>>> the "e", so the presence of the space is irrelevant.
>> That is not what I did.  I had the cursor over the space, in the
>> second column.  Maybe word breaks and/or column width has something to
>> do with this also.
> 
> Did you use the same recipe I gave in my OP, aside from the space?

Yes I did.

> With
> that and the space I still get the "empty" line, so as you say there
> must be some other difference.  When I do emacs -Q and eval
> (frame-parameters), I get this:
> 
> ((parent-id . 25170646)
>  (explicit-name)
>  (display . ":0")
>  (visibility . t)
>  (icon-name)
>  (outer-window-id . "65011885")
>  (window-id . "65011940")
>  (top . 30)
>  (left . 0)
>  (buried-buffer-list)
>  (buffer-list #<buffer *scratch*> #<buffer  *Minibuf-1*> #<buffer *Help*>)
>  (unsplittable)
>  (minibuffer . #<window 4 on  *Minibuf-0*>)
>  (modeline . t)
>  (width . 80)
>  (height . 40)
>  (name . "emacs@escher.local.home")
>  (environment)
>  (background-mode . light)
>  (display-type . color)
>  (horizontal-scroll-bars . t)
>  (window-system . x)
>  (scroll-bar-width . 16)
>  (cursor-type . box)
>  (auto-lower)
>  (auto-raise)
>  (icon-type . t)
>  (wait-for-wm . t)
>  (title)
>  (buffer-predicate)
>  (tool-bar-lines . 1)
>  (menu-bar-lines . 1)
>  (scroll-bar-background . "grey75")
>  (scroll-bar-foreground . "#000000")
>  (right-fringe . 8)
>  (left-fringe . 8)
>  (line-spacing)
>  (screen-gamma)
>  (border-color . "black")
>  (cursor-color . "#000000")
>  (mouse-color . "#000000")
>  (background-color . "#ffffff")
>  (foreground-color . "#000000")
>  (vertical-scroll-bars . left)
>  (internal-border-width . 0)
>  (border-width . 0)
>  (font . "-*-fixed-medium-r-*--16-*-*-*-*-*-fontset-16"))
> 
> Do you have different values for relevant parameters (e.g. font)?
> 

I have width 81, height 50 (from .Xresources) and font is 
"-*-*-medium-*-*-*-12-*-*-*-*-*-fontset-startup".

I also have right-fringe 11 and left-fringe 10.

	Jan D.

Re: 23.0.50; display property and continuation arrows

[Stephen Berman]

On Mon, 07 Jan 2008 08:00:06 +0100 Jan Djärv <jan.h.d@swipnet.se> wrote:

> Stephen Berman skrev:
>> On Sun, 06 Jan 2008 11:46:30 +0100 Jan Djärv <jan.h.d@swipnet.se> wrote:
>>
>>> Stephen Berman skrev:
>>>> On Sat, 05 Jan 2008 23:59:07 +0100 Jan Djärv <jan.h.d@swipnet.se> wrote:
>>>>
>>>>> Stephen Berman skrev:
>>>>>> On Sat, 05 Jan 2008 18:40:43 +0100 Jan Djärv <jan.h.d@swipnet.se> wrote:
>>>> [...]
>>>>>>> A futher observation.  If the character in the second column is
>>>>>>> a space, the bug does not appear.
>>>>>> I'm not sure what you mean: do you mean the position occupied by
>>>>>> "e" in the last occurrence of "testing" in my screen first shot?
>>>>>> If I make this a space, I see no different behavior.  Also, what
>>>>>> do you mean by the "bug"?
>>>>> The bug == the appearance of an "empty" line.  Yes, the e in your
>>>>> screen shot.
>>>> Ok, but in this case there is no "empty" line only if you apply the
>>>> display property with the cursor over the "s" following the space,
>>>> i.e. the cursor is on the third column.  But that's also so if you
>>>> leave the "e", so the presence of the space is irrelevant.
>>> That is not what I did.  I had the cursor over the space, in the
>>> second column.  Maybe word breaks and/or column width has something
>>> to do with this also.
>>
>> Did you use the same recipe I gave in my OP, aside from the space?
>
> Yes I did.
>
>> With that and the space I still get the "empty" line, so as you say
>> there must be some other difference.  When I do emacs -Q and eval
>> (frame-parameters), I get this:
[...]
>>  (width . 80)
>>  (height . 40)
[...]
>>  (right-fringe . 8)
>>  (left-fringe . 8)
[...]
>>  (font . "-*-fixed-medium-r-*--16-*-*-*-*-*-fontset-16"))
>>
>> Do you have different values for relevant parameters (e.g. font)?
>>
>
> I have width 81, height 50 (from .Xresources) and font is
> "-*-*-medium-*-*-*-12-*-*-*-*-*-fontset-startup".
>
> I also have right-fringe 11 and left-fringe 10.

When I start emacs -Q, then do (create-fontset-from-fontset-spec
"-*-*-medium-*-*-*-12-*-*-*-*-*-fontset-startup"), then (make-frame
'((width . 81) (height . 50) (right-fringe . 11) (left-fringe . 10)
(font . fontset-startup))), this results, with the text I used in my OP,
in the "e" in the last occurrence of "testing" being in the *first*
column of the continued line, not the second, as you reported -- it
cannot be otherwise with that text and frame width 81 instead of 80, at
least if the font used has a fixed width (in my case the font is
-Adobe-Courier-Medium-R-Normal--12-120-75-75-M-70-ISO8859-1).  But that
is the precisely position in which there is no "empty" line (nor
continuation arrows) after applying the display property, regardless of
whether the character is a space or not.  Are you sure you didn't apply
the display property to the first column instead the second?  Or are you
using a variable width font?

Steve Berman

Re: 23.0.50; display property and continuation arrows

[Jan Djärv]

Stephen Berman skrev:

> 
> When I start emacs -Q, then do (create-fontset-from-fontset-spec
> "-*-*-medium-*-*-*-12-*-*-*-*-*-fontset-startup"), then (make-frame
> '((width . 81) (height . 50) (right-fringe . 11) (left-fringe . 10)
> (font . fontset-startup))), this results, with the text I used in my OP,
> in the "e" in the last occurrence of "testing" being in the *first*
> column of the continued line, not the second, as you reported -- it
> cannot be otherwise with that text and frame width 81 instead of 80, at
> least if the font used has a fixed width (in my case the font is
> -Adobe-Courier-Medium-R-Normal--12-120-75-75-M-70-ISO8859-1).  But that
> is the precisely position in which there is no "empty" line (nor
> continuation arrows) after applying the display property, regardless of
> whether the character is a space or not.  Are you sure you didn't apply
> the display property to the first column instead the second?  Or are you
> using a variable width font?
> 

I'm not sure now what I did.  But I didn't use your "testing" line, your OP 
just said:

"2. Type a line of text that extends beyond (window-width) with auto-fill
mode is disabled so that continuation arrows are displayed in the
fringes."

I just typed in random characters.

But when I try now, I can't see any difference if the second character is a 
space or not.  Sorry for the noise.

	Jan D.

Re: 23.0.50; display property and continuation arrows

[Stephen Berman]

On Mon, 07 Jan 2008 11:42:53 +0100 Jan Djärv <jan.h.d@swipnet.se> wrote:

> Stephen Berman skrev:
>
>>
>> When I start emacs -Q, then do (create-fontset-from-fontset-spec
>> "-*-*-medium-*-*-*-12-*-*-*-*-*-fontset-startup"), then (make-frame
>> '((width . 81) (height . 50) (right-fringe . 11) (left-fringe . 10)
>> (font . fontset-startup))), this results, with the text I used in my OP,
>> in the "e" in the last occurrence of "testing" being in the *first*
>> column of the continued line, not the second, as you reported -- it
>> cannot be otherwise with that text and frame width 81 instead of 80, at
>> least if the font used has a fixed width (in my case the font is
>> -Adobe-Courier-Medium-R-Normal--12-120-75-75-M-70-ISO8859-1).  But that
>> is the precisely position in which there is no "empty" line (nor
>> continuation arrows) after applying the display property, regardless of
>> whether the character is a space or not.  Are you sure you didn't apply
>> the display property to the first column instead the second?  Or are you
>> using a variable width font?
>>
>
> I'm not sure now what I did.  But I didn't use your "testing" line,
> your OP just said:
>
> "2. Type a line of text that extends beyond (window-width) with auto-fill
> mode is disabled so that continuation arrows are displayed in the
> fringes."
>
> I just typed in random characters.
>
> But when I try now, I can't see any difference if the second character
> is a space or not.  Sorry for the noise.

No problem, at least we now know we observe the same behavior.  Since
Martin Rudalics also confirmed seeing it under w32, I wonder if RMS may
have tried something different from my recipe, since he reported not
getting the result without continuation arrows.  Richard, if you have
time could you try again and see if you really don't observe the same
thing?

Steve Berman