Closed GoogleCodeExporter closed 9 years ago
I've had the same problem, and I think it is related to whether or not you have
enabled extended input events. Also see:
https://bugs.launchpad.net/ubuntu/+source/cellwriter/+bug/290159
Original comment by Karl.Heg...@gmail.com
on 8 Feb 2010 at 2:01
See a possible workaround in issue 28. In my case, this sort of problem seemed
to be
caused by the introduction of windowless controls in GTK+, but compiling with
GNOME
solved it for the reporter of issue 28.
Original comment by bielaws...@gmail.com
on 9 Feb 2010 at 5:47
I get this problem with Archlinux. Setting the GDK_NATIVE_WINDOWS env var
doesn't help.
What actually happens is that you can do one stroke, and release the pen/mouse
button, and the stroke will not be cleared (unless you leave the box, of
course). However, as soon as you press down for the second stroke, it clears
the first one.
Original comment by goo...@randomguy3.me.uk
on 13 Jul 2010 at 2:25
Tell a lie - setting GDK_NATIVE_WINDOWS does fix it. I just hadn't quit
CellWriter properly before trying it.
Original comment by goo...@randomguy3.me.uk
on 13 Jul 2010 at 2:28
Bizarrely, if I compile with Gnome support, multi-stroke input works in normal
mode, but not training mode.
Original comment by goo...@randomguy3.me.uk
on 13 Jul 2010 at 2:30
Well, a bit of debugging shows that if GDK_NATIVE_WINDOWS is set, when I press
down with the stylus, I get a leave_notify_event with the mode set to
GDK_CROSSING_GRAB. This is ignored by cellwidget.c.
However, without GDK_NATIVE_WINDOWS set, the leave_notify_event has mode
GDK_CROSSING_NORMAL. No idea why yet.
Original comment by goo...@randomguy3.me.uk
on 13 Jul 2010 at 5:22
So, the spurious leave_notify event is triggered because accept_focus is set to
false on the window. This may well only happen with KWin - I haven't tried any
other window managers with my minimal test app.
The fact that it works with GTK_NATIVE_WINDOWS enabled (or GTK+ before 2.18.0)
appears to be a useful quirk of the cellwriter code. My minimal test app
(which has a gtk_drawing_area inside a window that outputs some debugging info
when it gets a leave_notify event) always produces GDK_CROSSING_NORMAL events
regardless of environment variables or anything else I've tried.
Note that simply calling gdk_window_ensure_native on the GdkWindow associated
with the cell_widget drawing_area is not sufficient (I tried calling
gdk_window_ensure_native(gtk_widget_get_window(widget)); in a realize event
handler). But GDK_NATIVE_WINDOWS apparently also changes some other behaviour,
according to
http://library.gnome.org/devel/gtk/2.18/gtk-migrating-ClientSideWindows.html.
Original comment by goo...@randomguy3.me.uk
on 13 Jul 2010 at 10:12
I can confirm that the XFCE4 window manager, as an example, doesn't produce
this behaviour.
Attached is a little test app.
Original comment by goo...@randomguy3.me.uk
on 14 Jul 2010 at 2:04
Attachments:
KWin bug filed (asking for more info):
http://bugs.kde.org/show_bug.cgi?id=244625
Original comment by goo...@randomguy3.me.uk
on 14 Jul 2010 at 4:25
Hi guys, sorry for letting this sit for so long. I will take a look at it over
the weekend. --risujin
Original comment by mele...@google.com
on 31 Jul 2010 at 4:16
I am running Kubuntu 11.04, and can confirm that launching with:
GDK_NATIVE_WINDOWS=1 cellwriter
fixed the single stroke problem for me also. Thank you so much. Is there a way
to make this happen by default for cellwriter. I don't mean mine, I mean the
program.
Original comment by SuperSma...@gmail.com
on 28 May 2011 at 7:35
Running cellwriter 1.3.4 on Ubuntu 11.04 x64 (Gnome classic) also seems to
produce this problem.
GDK_NATIVE_WINDOWS=1 cellwriter does appear to fix it, but this seems
temporary. Any solution?
Original comment by Zachary....@gmail.com
on 3 Jul 2011 at 3:59
still having this issue with 1.3.4 in ubuntu 11.04. any potential permanent
fixes?
Original comment by walil...@gmail.com
on 29 Sep 2011 at 5:08
Finally tested this with a Wacom tablet on a recent version of Ubuntu. It is
firing "leave_notify_event" like crazy, just removed that code and seems to
work well now.
Original comment by mele...@google.com
on 8 Oct 2011 at 6:31
Original issue reported on code.google.com by
andreasb...@gmail.com
on 14 Jan 2010 at 10:17