Help with scrolling - Emacs 22.3.1 windows

Re: Help with scrolling - Emacs 22.3.1 windows

On Friday, November 07, "Augusto Pedroza" <augusto.pedroza <at> gmail.com> wrote:
>I can scroll down with the mouse "wheel" in the first
>windows only even when I have selected the second
>window.

I am running an older version of emacs, but I, too,
observe somewhat surprising behaviour with mouse-wheel
scrolling in the presence of horizontally split
windows.  Once you have started scrolling in one
window, the same window will continue to scroll with
the mouse wheel independent of which window the mouse
pointer is hovering over.  However, if you stop
scrolling and wait for a short interval (about one
second), then, the next time you turn the scroll
wheel, the window then under the mouse pointer is the
one that will start scrolling.  In particular, the
window which scrolls is independent of which window is
selected.  

(Typically, if you click in a window to select it,
then, by the time you try to scroll with mouse pointer
still in same window, enough time will have passed
since you last scrolled that the newly selected window
is the one that will be scrolled.  I can get the
surprising behaviour after a click only if I really
work at it.)

Though this behaviour is somewhat surprising, I can
live with it.  Indeed, it may well have been a
conscious design choice.  However, I have failed to
find a parameter (e.g., the time interval after which
switching windows is possible) which might control
this behaviour.  The fact that the window reported as
being associated with a scroll event may not actually
change immediately when the mouse pointer moves into a
new window must result from something embedded down in
the low-level mouse handling routines.  It is not
reflected in the higher level code (w32-handle-scroll-
bar-event) which ultimately handles a given scrolling
event.

Regards,
  David V.

Re: Help with scrolling - Emacs 22.3.1 windows

David Vanderschel wrote:
> On Friday, November 07, "Augusto Pedroza" <augusto.pedroza <at> gmail.com> wrote:
>   
>> I can scroll down with the mouse "wheel" in the first
>> windows only even when I have selected the second
>> window.
>>     
>
> I am running an older version of emacs, but I, too,
> observe somewhat surprising behaviour with mouse-wheel
> scrolling in the presence of horizontally split
> windows.

What does C-h k [mouse-wheel] report?

Some mouse drivers do not simply send mouse-wheel events, but rather 
have their own (often buggy) code to locate the scroll bar and simulate 
scroll-bar events. If you do not use any special features of your mouse 
driver then you can try switching to one of the generic mouse drivers 
which will send proper wheel events.

Re: Help with scrolling - Emacs 22.3.1 windows

Re: Help with scrolling - Emacs 22.3.1 windows

On Friday, November 07, "Jason Rumney" <jasonr <at> f2s.com> wrote:
>David Vanderschel wrote:
>> On Friday, November 07, "Augusto Pedroza" <augusto.pedroza <at> gmail.com> wrote:

>>> I can scroll down with the mouse "wheel" in the first
>>> windows only even when I have selected the second
>>> window.

>> I am running an older version of emacs, but I, too,
>> observe somewhat surprising behaviour with mouse-wheel
>> scrolling in the presence of horizontally split
>> windows.

>What does C-h k [mouse-wheel] report?

Actually, regarding the code that might cause a switch
of windows affected by the scroll wheel, I had already
mentioned:
>>It is not reflected in the higher level code
>>(w32-handle-scroll- bar-event) which ultimately
>>handles a given scrolling event.

which does provide the answer to Jason's question.

>Some mouse drivers do not simply send mouse-wheel
>events, but rather have their own (often buggy) code
>to locate the scroll bar and simulate scroll-bar
>events.

Indeed.  I have such a 'mouse' also - a GlidePoint
touchpad on my keyboard.  However, the behaviour I was
reporting was for a Dell USB mouse which is using the
standard Windows USB pointing device driver.  With
some effort, I was also able to get an apparently
'wrong' behaviour with the touchpad as well; but its
behaviour is different with a difference more like I
suppose that Jason was considering.

>If you do not use any special features of your mouse
>driver then you can try switching to one of the
>generic mouse drivers which will send proper wheel
>events.

No.  I like the way both drivers behave.  What
strangeness exists is not a problem for me.  I said
that.  My point was that the symptom I was observing
might well be the same as Augusto was observing and
that he was possibly misinterpreting the symptom and
its seriousness.  (Indeed, his statement of the
symptom was hardly clear to start with.)  I was able
to observe a (not very problematic) behaviour which
could conceivably be described in the same manner as
he described the behaviour which was concerning him.
In particular, I tried to emphasize that which window
is selected is not relevant with respect to which
window scrolls and that this behaviour is clearly
intentional based on looking at the code.  

(The routine I mentioned in the quote above saves the
selected window and restores it after messing with the
scroll position of whichever window is associated with
the mouse event.  Being time-delay dependent, it is
that _association_ which behaves somewhat strangely,
and the responsibility for the strangeness may well
lie outside Emacs.  However, I do see time-delay
dependence with both drivers.)

It is conceivable that Augusto has a mouse driver
which is behaving completely wrongly with respect to
guessing which scroll bar should be associated with
the mouse wheel at a given point in time.  If that is
the case, then Jason's suggestion to try a simpler
driver could potentially solve what may be a real
problem for Augusto - but not for me.  Indeed, it may
be that Jason thought he was responding to Augusto
when, in fact, he was replying to my response to
Augusto.

Regards,
  David V.

Re: Help with scrolling - Emacs 22.3.1 windows

David Vanderschel wrote:

>  Indeed, it may be that Jason thought he was responding to Augusto
>  when, in fact, he was replying to my response to Augusto

I was responding to both of you, in particular since you said:

    It is not reflected in the higher level code (w32-handle-scroll-
    bar-event) which ultimately handles a given scrolling event.

Since w32-handle-scroll-bar-event is not involved in handling 
mouse-wheel events (at least in 22.1 and later, which uses generic mouse 
wheel code for all platforms).

CAPS LOCK does not work in cygwin shell - can't type uppercase - forced to use SHIFT

problems with RCS rcsdiff and checkout co.exe in emacs 22.3.1

Summary: In Emacs 22.3.1 on Windows XP some RCS utilities, e.g. checkin
ci.exe and checkout co.exe, do not work. If I cannot get RCS to work I
am open to suggestions about the simplest upgrade to another source
control system. 

Details:
I recently upgraded to GNU Emacs 22.3.1 (i386-mingw-nt5.1.2600) of
2008-09-06 on SOFT-MJASON . 
I was formerly running GNU Emacs 21
In the former emacs the RCS utilities all worked.  
I use an RCS directory (not the approach of having a ,v at the end of
the file name.)
All files are on my local PC running Windows XP.  There are no network
files involved.

In the new emacs 22.3.1 I have some problems with the RCS utilities.
Below is the simplest example.   I also have a problem doing a checkin
of an existing file after making an edit.  This problem is a
"showstopper" for me using the new emacs.
In emacs I create a new file named e.g. foo22, type "foo", and "C-x s"
to save the file, and then "C-x v v" to checkin.
That all works as expected.  The RCS/foo22 file is there and looks -OK.
(See below.)

Then I say "C-x v v" again to checkout, but it fails, with this message:
vc-do-command: Running rcsdiff...FAILED (status 2)

If I work in the DOS box (aka command window) I can do rcsdiff, co, and
ci without any problems.
I believe that rcsdiff.exe calls co.exe and something about co.exe is
the root of the problem.

I looked in the archives of both these mailing lists
help-emacs-windows <at> gnu.org and help-rcs <at> gnu.org without finding an
answer.

Here is the contents of the directory with the RCS utilities, plus a few
other related ones. 
I downloaded these years ago.  I think they are RCS version 5.7.  I
cannot find these on the web anymore. 

03/28/1999  05:07 PM            14,848 ci.exe
03/28/1999  05:12 PM            14,336 cmp.exe
03/28/1999  05:07 PM            11,264 co.exe
03/28/1999  05:12 PM            65,536 diff.exe
03/28/1999  05:12 PM            19,968 diff3.exe
03/28/1999  05:07 PM             4,608 ident.exe
03/28/1999  05:07 PM             4,608 merge.exe
11/13/2008  04:31 PM             2,693 o
03/28/1999  05:07 PM            15,872 rcs.exe
03/28/1999  05:07 PM             6,656 rcsclean.exe
03/28/1999  05:07 PM             7,680 rcsdiff.exe
03/28/1999  05:07 PM             7,680 rcsdiff_keep_older.exe
03/28/1999  05:07 PM            54,272 rcslib.dll
03/28/1999  05:07 PM             6,144 rcsmerge.exe
03/28/1999  05:07 PM            13,312 rlog.exe
03/28/1999  05:12 PM            17,408 sdiff.exe

Contents of a new file, here named foo22, just after it has been checked
in:

type RCS\foo22
head    1.1;
access;
symbols;
locks; strict;
comment  <at> #  <at> ;

1.1
date    2008.11.13.22.07.12;    author a071046; state Exp;
branches;
next    ;

desc
 <at>  <at> 

1.1
log
 <at> Initial revision
 <at> 
text
 <at> foo
 <at> 

Thanks in advance,
Steve

--
Steven Tolkin 
Vice President
PWI Strategy and Architecture 

Fidelity Investments
400 Puritan Way M3B
Marlborough MA 01752
tel: (508) 787-9006 
steve . . tolkin  <at>   <at>  fmr . . com

Notice:  All e-mail sent to or from Fidelity Investments is subject to
retention, monitoring and/or review by Fidelity personnel.
Please note that Fidelity is unable to accept orders left over voicemail
or email regarding any account.

The information in this e-mail and in any attachments is intended solely
for the attention and use of the named addressee(s) and may contain
information that is considered privileged, proprietary, confidential,
and/or exempt from disclosure under applicable law.  If you are not the
intended recipient of this email or if you have otherwise received this
email in error, please immediately notify me by replying to this message
or by telephone (you may call me collect).   Any use, dissemination,
distribution or copying of this e-mail is strictly prohibited without
authorization from Fidelity Investments.

Clearing, custody or other brokerage services may be provided by
Fidelity Brokerage Services LLC or National Financial Services LLC,
members NYSE, SIPC.

Re: problems with RCS rcsdiff and checkout co.exe in emacs 22.3.1

> Date: Thu, 13 Nov 2008 17:26:13 -0500
> From: "Tolkin, Steve" <Steve.Tolkin <at> FMR.COM>
> Cc: help-rcs <at> gnu.org
> 
> In the new emacs 22.3.1 I have some problems with the RCS utilities.
> Below is the simplest example.   I also have a problem doing a checkin
> of an existing file after making an edit.  This problem is a
> "showstopper" for me using the new emacs.
> In emacs I create a new file named e.g. foo22, type "foo", and "C-x s"
> to save the file, and then "C-x v v" to checkin.
> That all works as expected.  The RCS/foo22 file is there and looks -OK.
> (See below.)
> 
> Then I say "C-x v v" again to checkout, but it fails, with this message:
> vc-do-command: Running rcsdiff...FAILED (status 2)

I cannot reproduce this on my machine, using Emacs 22.3 invoked with
"emacs -Q".  If you use "emacs -Q", does the problem persist?

If "emacs -Q" doesn't work either, then please tell where did you get
the ported RCS package you are using.  Perhaps that port is somehow
subtly incompatible with how Emacs invokes subsidiary programs.  (I
needed to roll my own port of RCS because the one I downloaded from
GnuWin32 was buggy, and rcsdiff was indeed one of the problems.
However, in my case, it didn't work even from the command line.)

> I use an RCS directory (not the approach of having a ,v at the end of
> the file name.)

But the files inside the RCS/ subdirectory do have the ,v suffix,
right?

23.0.60; vc-git and tramp interference on EmacsW32

When using Tramp connections to a remote unix system, with vc-git 
enabled, while browsing in the source
tree of a git repository, the use of 'git ls-files' produces strange, 
unwanted, behaviour that interferes
with the functioning of ido-mode and possibly others.

Specifically, once vc-git has identified that the remote directory is 
part of a git repository, it
starts using 'git ls-files -c -z --' to obtain file information instead 
of the normal method of ls/stat.
Unfortunately, every filename that is returned this way seems to have a 
^M appended onto the end of
each filename and directory name, confusing the ssh connection into 
appending ^M onto every line,
breaking and subsequently ido-mode's file browsing.

Examples from tramp debug:

Prior to loading a git maintained file - all systems normal :

17:04:58 tramp-send-string (10) # /bin/ls -ab 2>/dev/null | while read 
f; do if test -d "$f" 2>/dev/null; then echo "$f/"; else echo "$f"; fi; done
17:04:58 tramp-get-connection-property (7) # process-buffer nil
17:04:58 tramp-get-connection-property (7) # check-remote-echo nil
17:04:58 tramp-get-connection-property (7) # check-remote-echo nil
17:04:58 tramp-accept-process-output (10) # *tramp/plink nikki <at> cthulu* run
17:04:58 tramp-accept-process-output (10) #
./
../
Main.cpp
Main.cpp.~1~
Main.o
///b5a78361da09334ffa171b88de21fa05

After loading Main.cpp into a buffer, and thus invoking vc-git's control 
over the situation:

17:05:02 tramp-send-string (10) # git ls-files -c -z -- Main.cpp 
</dev/null 2>/dev/null
17:05:02 tramp-get-connection-property (7) # process-buffer nil
17:05:02 tramp-get-connection-property (7) # check-remote-echo nil
17:05:02 tramp-get-connection-property (7) # check-remote-echo nil
17:05:02 tramp-accept-process-output (10) # *tramp/plink nikki <at> cthulu* run
17:05:02 tramp-accept-process-output (10) #
Main.cpp^M
///b5a78361da09334ffa171b88de21fa05^M^ <at> 

17:05:06 tramp-send-string (10) # /bin/ls -ab 2>/dev/null | while read 
f; do if test -d "$f" 2>/dev/null; then echo "$f/"; else echo "$f"; fi; done
17:05:06 tramp-get-connection-property (7) # process-buffer nil
17:05:06 tramp-get-connection-property (7) # check-remote-echo nil
17:05:06 tramp-get-connection-property (7) # check-remote-echo nil
17:05:06 tramp-accept-process-output (10) # *tramp/plink nikki <at> cthulu* run
17:05:06 tramp-accept-process-output (10) #
./^M
../^M
Main.cpp^M
Main.cpp.~1~^M
Main.o^M
^M
///b5a78361da09334ffa171b88de21fa05^M^ <at> 

I'm unsure if this is a bug in vc-git, tramp, emacsw32, or 'git 
ls-files', but I imagine that filtering spurious ^Ms in tramp is 
probably the easiest workaround.

In GNU Emacs 23.0.60.1 (i386-mingw-nt5.1.2600)
 of 2008-09-09 on LENNART-69DE564 (patched)
Windowing system distributor `Microsoft Corp.', version 5.1.2600
configured using `configure --with-gcc (3.4) --cflags -Ic:/g/include'

Important settings:
  value of $LC_ALL: nil
  value of $LC_COLLATE: nil
  value of $LC_CTYPE: nil
  value of $LC_MESSAGES: nil
  value of $LC_MONETARY: nil
  value of $LC_NUMERIC: nil
  value of $LC_TIME: nil
  value of $LANG: ENG
  value of $XMODIFIERS: nil
  locale-coding-system: cp1252
  default-enable-multibyte-characters: t

Major mode: C++/l

Minor modes in effect:
  highlight-changes-visible-mode: t
  shell-dirtrack-mode: t
  tooltip-mode: t
  mouse-wheel-mode: t
  noticeable-minibuffer-prompts-mode: t
  menu-bar-mode: t
  file-name-shadow-mode: t
  global-font-lock-mode: t
  font-lock-mode: t
  blink-cursor-mode: t
  global-auto-composition-mode: t
  auto-composition-mode: t
  auto-encryption-mode: t
  auto-compression-mode: t
  column-number-mode: t
  line-number-mode: t
  transient-mark-mode: t
  abbrev-mode: t

Recent input:
SPC <backspace> ( S-SPC ^ <S-backspace> & m s g S-SPC
) ; C-x C-s <up> <up> <up> <up> <up> <down> <down>
<down> <up> <up> <up> <up> <up> <up> <up> <up> <up>
<up> <up> <up> <up> <up> <up> <up> <up> <up> <up> <up>
<up> <up> <up> <up> <up> <up> <up> <up> <up> <up> <up>
<up> <up> C-x o <C-home> <next> <next> <next> <next>
<next> <prior> <next> <next> <next> C-x o L R E S U
L T S-SPC C A L L B A C K SPC M a i n W i n d o w C
l a l <backspace> <backspace> <backspace> M-/ SPC <backspace>
( SPC H W N D SPC h W n d , SPC W O R D SPC m s g ,
SPC W O R D SPC w P a r a m SPC <backspace> , <return>
<tab> D W O R D SPC l P a r a m SPC ) SPC { <return>
} <up> <up> <end> <return> <backspace> <down> <return>
<return> <tab> r e t u r n SPC 0 ; <up> C-x o <C-down>
<C-down> <up> <up> <up> <up> C-SPC <C-down> <up> <down>
<down> <down> M-w C-x o C-y <up> <up> <up> <up> <up>
<up> <return> <tab> <down> <tab> <down> <tab> <down>
<tab> <down> <tab> <down> <tab> <down> <tab> <down>
<tab> <up> <up> <up> <up> <up> <up> <up> <up> <down>
<down> C-x C-s <C-end> <up> <prior> <down> <down> <down>
<down> <up> <return> <tab> <help-echo> M-x m o d i
e <tab> <backspace> <backspace> <backspace> <backspace>
<backspace> m e n u <tab> - b a r <tab> m o d e <return>
<help-echo> <help-echo> <help-echo> <help-echo> <help-echo>
<help-echo> <help-echo> <help-echo> <help-echo> <help-echo>
<help-echo> <help-echo> <help-echo> <help-echo> <help-echo>
<help-echo> <help-echo> <help-echo> <help-echo> <help-echo>
<help-echo> <help-echo> <help-echo> <help-echo> <menu-bar>
<help-menu> <send-emacs-bug-report>

Recent messages:
Auto-saving...done
Making completion list...
Menu-Bar mode enabled