keyboard error (the strangest thing...)

[Brett Jones]

Still hoping for a little help...

I've verified that vncviewer is sending what looks to be the correct key codes
by using the logging functionality.  Here's the log (level 8) for pressing 4
thru 6 on the regular number keys:

 keyData 50001 not in special keymap, ToAscii (without ctrl) returns 1
byte(s): 34 (4) SendKeyEvent: key = x0034 status = down
Sent keysym 0034 (press)
 keyData c0050001 not in special keymap, ToAscii (without ctrl) returns 1
byte(s): 34 (4) SendKeyEvent: key = x0034 status = up
Sent keysym 0034 (release)
 keyData 60001 not in special keymap, ToAscii (without ctrl) returns 1
byte(s): 35 (5) SendKeyEvent: key = x0035 status = down
Sent keysym 0035 (press)
 keyData c0060001 not in special keymap, ToAscii (without ctrl) returns 1
byte(s): 35 (5) SendKeyEvent: key = x0035 status = up
Sent keysym 0035 (release)
 keyData 70001 not in special keymap, ToAscii (without ctrl) returns 1
byte(s): 36 (6) SendKeyEvent: key = x0036 status = down
Sent keysym 0036 (press)
 keyData c0070001 not in special keymap, ToAscii (without ctrl) returns 1
byte(s): 36 (6) SendKeyEvent: key = x0036 status = up
Sent keysym 0036 (release)

I don't notice anything different about the 5 key.  Here's the 5 key, followed
by the numeric keyboard 5 key:

 keyData 60001 not in special keymap, ToAscii (without ctrl) returns 1
byte(s): 35 (5) SendKeyEvent: key = x0035 status = down
Sent keysym 0035 (press)
 keyData c0060001 not in special keymap, ToAscii (without ctrl) returns 1
byte(s): 35 (5) SendKeyEvent: key = x0035 status = up
Sent keysym 0035 (release)
 keyData 4c0001 keymap gives 65461 (ffb5) SendKeyEvent: key = xffb5 status =
down
Sent keysym ffb5 (press)
 keyData c04c0001 keymap gives 65461 (ffb5) SendKeyEvent: key = xffb5 status =
up
Sent keysym ffb5 (release)

just for kicks, showing that indeed there are different codes for the two
different keys.


So... the problem must be on the server side.  (This jives with the fact that
the mystery 5 problem doesn't happen on a WinVNC server.)  I got the source
and built it myself (redhat 6.0, 2.2.12 kernel) but the problem still is out
there.

Can anyone tell me if there is a similar logging functionality in the server
code?  The size of the code is daunting and I can't find my way around in it
too easily.  I guess I'm wondering:

- is there a logging functionality already in place on the server side like
the client side?
- where in the server code would I look for where these keysym events arrive?
- and where are they translated back into a nice little ascii value?

Thanks for any help!

Brett




Brett Jones wrote:

> I have VNC running on several machines, but usually use it to view a
> linux server on an NT client.  I've been using it this way for quite
> some time, but just recently upgraded to RedHat 6.0.  At some point near
> or shortly after the upgrade, I started getting the strangest error.
>
> Any time I hit the '5' key (on the NT keyboard), I get a backspace
> inside VNC.  The 5 key is the only key this happens with.  I cannot get
> a percent sign now, as pressing 5 with either shift key still remains a
> backspace.  I'm assuming it's backspace, because pressing ^V and then 5
> yields the same control code on the screen as ^V Backspace.  (Normally
> it just erases text.)
>
> - The 5 key is the only key not working.
> - I've tried it from several different windoze machines for viewers (NT
> and 98).
> - Pressing the 5 key on the linux console (while in X) works correctly.
> - The same effect happens regardless of linux user running vncserver.
> - I'm using vnc-3.3.3-3, which I downloaded from the VNC site in an RPM
> - These are just plain old fashioned American keyboards, not
> international
> - I'm really not exactly sure if it stopped working right after I moved
> to RH 6.0, or if it happed some time after that.  It took me a while to
> figure out what the flip was going on.
>
> I'm in a semi-workable state, since I can use the num-key 5 when I need
> it, but I'm going to need a percent sign at some point in the near
> future, so I've got to get this figured out.
>
> Has anyone heard about something like this?
>
> Thanks,
>
> Brett Jones
>
> ---------------------------------------------------------------------
> The VNC mailing list - see http://www.uk.research.att.com/vnc/intouch.html
> ---------------------------------------------------------------------


---------------------------------------------------------------------
The VNC mailing list - see http://www.uk.research.att.com/vnc/intouch.html
---------------------------------------------------------------------