I got a new monitor recently and started to work with
two screens and dwm.
No problem at all. Everything works fine, except for one
_really annoying_ issue with dmenu and dwm.
If no windows are open on the current tag of one screen
and I switch to this screen&tag using the keyboard
then if I invoke dmenu it doesn't always appear on the focused
screen, but on the screen where the cursor currently is!
The issue is gone if I open at least one client on the focused
screen before invoking dmenu.
I've got no time at the moment to investigate this and don't even know
where to start looking. Could someone who knows a bit more about
X internals than me please investigate this problem!
On Tue, Aug 14, 2012 at 9:01 AM, v4hn <m...@v4hn.de> wrote:
> Hey everyone,
> I got a new monitor recently and started to work with
> two screens and dwm.
> No problem at all. Everything works fine, except for one
> _really annoying_ issue with dmenu and dwm.
> If no windows are open on the current tag of one screen
> and I switch to this screen&tag using the keyboard
> then if I invoke dmenu it doesn't always appear on the focused
> screen, but on the screen where the cursor currently is!
> The issue is gone if I open at least one client on the focused
> screen before invoking dmenu.
I started using warp because moving my mouse from one screen to another
manually was a pain in the arse. That might fix your issue.
--Carlos
> I've got no time at the moment to investigate this and don't even know
> where to start looking. Could someone who knows a bit more about
> X internals than me please investigate this problem!
On Tue, Aug 14, 2012 at 03:01:08PM +0200, v4hn wrote:
> Hey everyone,
> I got a new monitor recently and started to work with
> two screens and dwm.
> No problem at all. Everything works fine, except for one
> _really annoying_ issue with dmenu and dwm.
> If no windows are open on the current tag of one screen
> and I switch to this screen&tag using the keyboard
> then if I invoke dmenu it doesn't always appear on the focused
> screen, but on the screen where the cursor currently is!
> The issue is gone if I open at least one client on the focused
> screen before invoking dmenu.
> I've got no time at the moment to investigate this and don't even know
> where to start looking. Could someone who knows a bit more about
> X internals than me please investigate this problem!
> v4hn
[1] for dwm, [2] for dmenu, accordingly.
I did publish it before, no idea where it got lost.
Otoh, it's a fairly simple addition, which to figure out isn't much to imo.
On Tue, Aug 14, 2012 at 9:33 AM, Martti Kühne <mysat...@gmail.com> wrote:
> On Tue, Aug 14, 2012 at 03:01:08PM +0200, v4hn wrote:
> > Hey everyone,
> > I got a new monitor recently and started to work with
> > two screens and dwm.
> > No problem at all. Everything works fine, except for one
> > _really annoying_ issue with dmenu and dwm.
> > If no windows are open on the current tag of one screen
> > and I switch to this screen&tag using the keyboard
> > then if I invoke dmenu it doesn't always appear on the focused
> > screen, but on the screen where the cursor currently is!
> > The issue is gone if I open at least one client on the focused
> > screen before invoking dmenu.
> > I've got no time at the moment to investigate this and don't even know
> > where to start looking. Could someone who knows a bit more about
> > X internals than me please investigate this problem!
> > v4hn
> [1] for dwm, [2] for dmenu, accordingly.
> I did publish it before, no idea where it got lost.
> Otoh, it's a fairly simple addition, which to figure out isn't much to
> imo.
pass NULL in config.h keys[], as an arg, it's needed to provide a common ground
call structure for all keyboard functions. Also, line 32 of the dwm patch might
be written shorter with +=, since it's initialized correctly in 16
these patches work like a charme, thanks!
If this is the most simple way of fixing this mess,
then I vote for adding it to upstream.
Still this looks like patchwork to me..
dmenu fails to detect the active screen in dwm.
So either the detection is broken or dwm does not support
the detection used. Is there a standardized way of doing this?
Could the authors of the respective code lines please have a look at this?
On Tue, Aug 14, 2012 at 03:33:45PM +0200, Martti Kühne wrote:
> [1] for dwm, [2] for dmenu, accordingly.
> I did publish it before, no idea where it got lost.
> Otoh, it's a fairly simple addition, which to figure out isn't much to imo.
On Thu, Aug 16, 2012 at 02:14:53PM +0200, v4hn wrote:
> Hey Martti,
> these patches work like a charme, thanks!
> If this is the most simple way of fixing this mess,
> then I vote for adding it to upstream.
> Still this looks like patchwork to me..
> dmenu fails to detect the active screen in dwm.
> So either the detection is broken or dwm does not support
> the detection used. Is there a standardized way of doing this?
> Could the authors of the respective code lines please have a look at this?
Urm, afaik I am the author, and I wrote it for lack a better alternative. If
there would be the whole integration would be easier.
If I remember it correctly, X can tell what monitor the mouse is on, but that
state was made invalid when a key combo to switch monitor was introduced.
Don't know, maybe someone else can clarify this more thoroughly or you could
read into Xlib docs or code from other (fully keyboard controllable) wms to see
if/how they do it right...
Also the patch against dwm I provided about ewmh is still not working as it
should, but I wasn't annoyed enough yet to write a patch to dump the whole ewmh
murks. Ironically that one was actually added upstream.
