bug#687: 23.0.60; Mouse wheel should not trigger mouse avoidance

bug#687: 23.0.60; Mouse wheel should not trigger mouse avoidance

[Lennart Borgman (gmail)]

[-- Attachment #1: Type: text/plain, Size: 473 bytes --]

I noticed that mouse wheel events might trigger mouse avoidance. That
seems inappropriate. The attached patch should fix that.

Please notice that I am a bit unsure about what the event names could
be. There are probably more event names than those I have entered. In
keyboard.c there is the following code:

static char *lispy_wheel_names[] =
{
   "wheel-up", "wheel-down", "wheel-left", "wheel-right"
};



In GNU Emacs 23.0.60.1 (i386-mingw-nt5.1.2600)
  of 2008-08-10


[-- Attachment #2: avoid-wheel.diff --]
[-- Type: text/plain, Size: 710 bytes --]

Index: avoid.el
===================================================================
RCS file: /sources/emacs/emacs/lisp/avoid.el,v
retrieving revision 1.47
diff -u -r1.47 avoid.el
--- avoid.el	6 May 2008 07:57:26 -0000	1.47
+++ avoid.el	10 Aug 2008 20:42:26 -0000
@@ -289,7 +289,9 @@
 	     (let ((modifiers (event-modifiers (car last-input-event))))
 	       (or (memq (car last-input-event)
 			 '(mouse-movement scroll-bar-movement
-			   select-window switch-frame))
+			   select-window switch-frame
+                           wheel-down wheel-up double-wheel-up double-wheel-down
+                           ))
 		   (memq 'click modifiers)
 		   (memq 'double modifiers)
 		   (memq 'triple modifiers)

bug#687: 23.0.60; Mouse wheel should not trigger mouse avoidance

[ＯＦＦＩＣＥ ZERO]

Do not send me any mail  !
no thank you



----- Original Message ----- 
From: "Lennart Borgman (gmail)" <lennart.borgman@gmail.com>
To: <emacs-pretest-bug@gnu.org>
Sent: Monday, August 11, 2008 5:58 AM
Subject: bug#687: 23.0.60; Mouse wheel should not trigger mouse avoidance


>I noticed that mouse wheel events might trigger mouse avoidance. That
> seems inappropriate. The attached patch should fix that.
>
> Please notice that I am a bit unsure about what the event names could
> be. There are probably more event names than those I have entered. In
> keyboard.c there is the following code:
>
> static char *lispy_wheel_names[] =
> {
>   "wheel-up", "wheel-down", "wheel-left", "wheel-right"
> };
>
>
>
> In GNU Emacs 23.0.60.1 (i386-mingw-nt5.1.2600)
>  of 2008-08-10
>
>


--------------------------------------------------------------------------------


> Index: avoid.el
> ===================================================================
> RCS file: /sources/emacs/emacs/lisp/avoid.el,v
> retrieving revision 1.47
> diff -u -r1.47 avoid.el
> --- avoid.el 6 May 2008 07:57:26 -0000 1.47
> +++ avoid.el 10 Aug 2008 20:42:26 -0000
> @@ -289,7 +289,9 @@
>       (let ((modifiers (event-modifiers (car last-input-event))))
>         (or (memq (car last-input-event)
>  '(mouse-movement scroll-bar-movement
> -    select-window switch-frame))
> +    select-window switch-frame
> +                           wheel-down wheel-up double-wheel-up 
> double-wheel-down
> +                           ))
>     (memq 'click modifiers)
>     (memq 'double modifiers)
>     (memq 'triple modifiers)
> 

bug#687: marked as done (23.0.60; Mouse wheel should not trigger mouse avoidance)

[Emacs bug Tracking System]

[-- Attachment #1: Type: text/plain, Size: 900 bytes --]


Your message dated Sat, 03 Jan 2009 23:02:09 +0800
with message-id <495F7DF1.9000705@gnu.org>
and subject line Re: bug#687: 23.0.60; Mouse wheel should not trigger mouse avoidance
has caused the Emacs bug report #687,
regarding 23.0.60; Mouse wheel should not trigger mouse avoidance
to be marked as done.

This means that you claim that the problem has been dealt with.
If this is not the case it is now your responsibility to reopen the
bug report if necessary, and/or fix the problem forthwith.

(NB: If you are a system administrator and have no idea what this
message is talking about, this may indicate a serious mail system
misconfiguration somewhere. Please contact owner@emacsbugs.donarmstrong.com
immediately.)


-- 
687: http://emacsbugs.donarmstrong.com/cgi-bin/bugreport.cgi?bug=687
Emacs Bug Tracking System
Contact owner@emacsbugs.donarmstrong.com with problems

[-- Attachment #2: Type: message/rfc822, Size: 3969 bytes --]

[-- Attachment #2.1.1: Type: text/plain, Size: 473 bytes --]

I noticed that mouse wheel events might trigger mouse avoidance. That
seems inappropriate. The attached patch should fix that.

Please notice that I am a bit unsure about what the event names could
be. There are probably more event names than those I have entered. In
keyboard.c there is the following code:

static char *lispy_wheel_names[] =
{
   "wheel-up", "wheel-down", "wheel-left", "wheel-right"
};



In GNU Emacs 23.0.60.1 (i386-mingw-nt5.1.2600)
  of 2008-08-10


[-- Attachment #2.1.2: avoid-wheel.diff --]
[-- Type: text/plain, Size: 710 bytes --]

Index: avoid.el
===================================================================
RCS file: /sources/emacs/emacs/lisp/avoid.el,v
retrieving revision 1.47
diff -u -r1.47 avoid.el
--- avoid.el	6 May 2008 07:57:26 -0000	1.47
+++ avoid.el	10 Aug 2008 20:42:26 -0000
@@ -289,7 +289,9 @@
 	     (let ((modifiers (event-modifiers (car last-input-event))))
 	       (or (memq (car last-input-event)
 			 '(mouse-movement scroll-bar-movement
-			   select-window switch-frame))
+			   select-window switch-frame
+                           wheel-down wheel-up double-wheel-up double-wheel-down
+                           ))
 		   (memq 'click modifiers)
 		   (memq 'double modifiers)
 		   (memq 'triple modifiers)

[-- Attachment #3: Type: message/rfc822, Size: 3844 bytes --]

From: Jason Rumney <jasonr@gnu.org>
To: 687-done@emacsbugs.donarmstrong.com
Cc: "Lennart Borgman (gmail)" <lennart.borgman@gmail.com>, Chong Yidong <cyd@stupidchicken.com>
Subject: Re: bug#687: 23.0.60; Mouse wheel should not trigger mouse avoidance
Date: Sat, 03 Jan 2009 23:02:09 +0800
Message-ID: <495F7DF1.9000705@gnu.org>

Jason Rumney wrote:
> Chong Yidong wrote:
>> I see.  Can you verify that Lennart's patch works on W32?
>>   
>
> Lennart's patch works, but it might be better to make wheel events on 
> W32 set the appropriate click, double or triple modifier so that it is 
> caught by the conditions that follow.

I have changed the code to make wheel events appear as click events 
(double and triple wheel events were already handled correctly).

I am still confused about all the steps an event goes through to become 
a Lisp event, as the click modifier was being set on the wheel events, 
but wasn't returned by event-modifiers, which seems to ignore the 
modifiers and just use the base name of the event to regenerate them. 
Are the modifiers we set in C code that untrustworthy?

bug#687: 23.0.60; Mouse wheel should not trigger mouse avoidance

[Chong Yidong]

> I noticed that mouse wheel events might trigger mouse avoidance. That
> seems inappropriate. The attached patch should fix that.

Doesn't the code just below your change prevent mouse wheel events from
triggering avoidance?

   (memq 'click modifiers)
   ...
   (memq 'down modifiers)))))))

Do you have a recipe for reproducing the bug?

bug#687: 23.0.60; Mouse wheel should not trigger mouse avoidance

[Lennart Borgman (gmail)]

Chong Yidong wrote:
>> I noticed that mouse wheel events might trigger mouse avoidance. That
>> seems inappropriate. The attached patch should fix that.
> 
> Doesn't the code just below your change prevent mouse wheel events from
> triggering avoidance?
> 
>    (memq 'click modifiers)
>    ...
>    (memq 'down modifiers)))))))
> 
> Do you have a recipe for reproducing the bug?

Not sure, but my fix does not cure the problem. I think I understand 
better how to reproduce it now:

- Move the mouse pointer while scrolling with the scroll wheel.

Can you reproduce the problem this way?

bug#687: 23.0.60; Mouse wheel should not trigger mouse avoidance

[Chong Yidong]

"Lennart Borgman (gmail)" <lennart.borgman@gmail.com> writes:

> Not sure, but my fix does not cure the problem. I think I understand
> better how to reproduce it now:
>
> - Move the mouse pointer while scrolling with the scroll wheel.
>
> Can you reproduce the problem this way?

I don't see anything out of the ordinary.  You're going to have to be
more specific.

bug#687: 23.0.60; Mouse wheel should not trigger mouse avoidance

[Lennart Borgman (gmail)]

Chong Yidong wrote:
> "Lennart Borgman (gmail)" <lennart.borgman@gmail.com> writes:
> 
>> Not sure, but my fix does not cure the problem. I think I understand
>> better how to reproduce it now:
>>
>> - Move the mouse pointer while scrolling with the scroll wheel.
>>
>> Can you reproduce the problem this way?
> 
> I don't see anything out of the ordinary.  You're going to have to be
> more specific.

Ok:

   emacs -Q
   M-x mouse-avoidance-mode
   Select ...: banish
   C-x 2

Now use the scroll wheel to try to scroll the buffers while 
simultaneously moving the mouse pointer.

bug#687: 23.0.60; Mouse wheel should not trigger mouse avoidance

[Chong Yidong]

"Lennart Borgman (gmail)" <lennart.borgman@gmail.com> writes:

>>> - Move the mouse pointer while scrolling with the scroll wheel.
>>>
>>> Can you reproduce the problem this way?
>>
>> I don't see anything out of the ordinary.  You're going to have to be
>> more specific.
>
> Ok:
>
>   emacs -Q
>   M-x mouse-avoidance-mode
>   Select ...: banish
>   C-x 2
>
> Now use the scroll wheel to try to scroll the buffers while
> simultaneously moving the mouse pointer.

I don't observe anything unusual when I do this.  What am I supposed to
see?

bug#687: 23.0.60; Mouse wheel should not trigger mouse avoidance

[Lennart Borgman (gmail)]

Chong Yidong wrote:
> "Lennart Borgman (gmail)" <lennart.borgman@gmail.com> writes:
> 
>>>> - Move the mouse pointer while scrolling with the scroll wheel.
>>>>
>>>> Can you reproduce the problem this way?
>>> I don't see anything out of the ordinary.  You're going to have to be
>>> more specific.
>> Ok:
>>
>>   emacs -Q
>>   M-x mouse-avoidance-mode
>>   Select ...: banish
>>   C-x 2
>>
>> Now use the scroll wheel to try to scroll the buffers while
>> simultaneously moving the mouse pointer.
> 
> I don't observe anything unusual when I do this.  What am I supposed to
> see?

For me the mouse pointer suddenly jumps away to its "bannished corner". 
Does not that happen for you?

bug#687: 23.0.60; Mouse wheel should not trigger mouse avoidance

[Chong Yidong]

"Lennart Borgman (gmail)" <lennart.borgman@gmail.com> writes:

> For me the mouse pointer suddenly jumps away to its "bannished
> corner". Does not that happen for you?

No.

This may be a W32 bug; maybe someone on that platform can try to
reproduce it.

bug#687: 23.0.60; Mouse wheel should not trigger mouse avoidance

[Jason Rumney]

Chong Yidong wrote:
> "Lennart Borgman (gmail)" <lennart.borgman@gmail.com> writes:
>
>   
>> For me the mouse pointer suddenly jumps away to its "bannished
>> corner". Does not that happen for you?
>>     
>
> No.
>
> This may be a W32 bug; maybe someone on that platform can try to
> reproduce it.
>   

This is probably because on W32 the mouse wheel generates wheel-up and 
wheel-down events, while on X it most commonly generates mouse-4 and 
mouse-5 button events. Whatever code is filtering events from affecting 
mouse-avoidance-mode is not filtering out the wheel events.

bug#687: 23.0.60; Mouse wheel should not trigger mouse avoidance

[Chong Yidong]

Jason Rumney <jasonr@f2s.com> writes:

> Chong Yidong wrote:
>> "Lennart Borgman (gmail)" <lennart.borgman@gmail.com> writes:
>>
>>   
>>> For me the mouse pointer suddenly jumps away to its "bannished
>>> corner". Does not that happen for you?
>>>     
>>
>> No.
>>
>> This may be a W32 bug; maybe someone on that platform can try to
>> reproduce it.
>>   
>
> This is probably because on W32 the mouse wheel generates wheel-up and
> wheel-down events, while on X it most commonly generates mouse-4 and
> mouse-5 button events. Whatever code is filtering events from
> affecting mouse-avoidance-mode is not filtering out the wheel events.

I see.  Can you verify that Lennart's patch works on W32?

bug#687: 23.0.60; Mouse wheel should not trigger mouse avoidance

[Jason Rumney]

Chong Yidong wrote:
> I see.  Can you verify that Lennart's patch works on W32?
>   

Lennart's patch works, but it might be better to make wheel events on 
W32 set the appropriate click, double or triple modifier so that it is 
caught by the conditions that follow.

bug#687: 23.0.60; Mouse wheel should not trigger mouse avoidance

[Lennart Borgman (gmail)]

Jason Rumney wrote:
> Chong Yidong wrote:
>> "Lennart Borgman (gmail)" <lennart.borgman@gmail.com> writes:
>>
>>  
>>> For me the mouse pointer suddenly jumps away to its "bannished
>>> corner". Does not that happen for you?
>>>     
>>
>> No.
>>
>> This may be a W32 bug; maybe someone on that platform can try to
>> reproduce it.
>>   
> 
> This is probably because on W32 the mouse wheel generates wheel-up and 
> wheel-down events, while on X it most commonly generates mouse-4 and 
> mouse-5 button events. Whatever code is filtering events from affecting 
> mouse-avoidance-mode is not filtering out the wheel events.

I just tested again with a (patched) checkout from 2008-08-13. I can not 
reproduce the problem there. (But it is still there when I check with 
the unpatched version from 2008-08-10.)

Has anyone found a solution, or?

bug#687: 23.0.60; Mouse wheel should not trigger mouse avoidance

[Lennart Borgman (gmail)]

Lennart Borgman (gmail) wrote:
> Jason Rumney wrote:
>> Chong Yidong wrote:
>>> "Lennart Borgman (gmail)" <lennart.borgman@gmail.com> writes:
>>>
>>>  
>>>> For me the mouse pointer suddenly jumps away to its "bannished
>>>> corner". Does not that happen for you?
>>>>     
>>>
>>> No.
>>>
>>> This may be a W32 bug; maybe someone on that platform can try to
>>> reproduce it.
>>>   
>>
>> This is probably because on W32 the mouse wheel generates wheel-up and 
>> wheel-down events, while on X it most commonly generates mouse-4 and 
>> mouse-5 button events. Whatever code is filtering events from 
>> affecting mouse-avoidance-mode is not filtering out the wheel events.
> 
> I just tested again with a (patched) checkout from 2008-08-13. I can not 
> reproduce the problem there. (But it is still there when I check with 
> the unpatched version from 2008-08-10.)
> 
> Has anyone found a solution, or?


Maybe it is because I am actually using the patch I sent ...

I hope you have get used to my temporary confusions.