05:30 pavlushka: issues with copy pasting, how do I do it?, copy pasting between gui windows
09:58 __monty__: Does libinput debug-events bypass the configuration just as debug-gui does? I'm trying to troubleshoot why middle click won't close Firefox tabs for me. Middle drag works for OpenBox but middle click doesn't. Yet debug-events does show middle click events.
10:02 vyivel: __monty__: device configuration is per-context, yes
10:03 vyivel: my guess is a compositor-specific setting or a bug, try with some other compositor to confirm
10:03 __monty__: Documentation for debug-gui says to pass flags like `--enable-tap` to debug your user configuration but my debug-events does not accept such flags. Or is that not what you mean when you say context?
10:06 vyivel: i mean that device configuration is applied within a specific program
10:06 vyivel: like, a compositor is a libinput user, it supplies device configuration from its settings or whatever
10:07 vyivel: "libinput debug-events" is another libinput user, it doesn't supply device configuration and therefore gets different results from same devices
10:07 __monty__: Wait, are we talking Wayland compositor? Given the channel I understand the assumption but I'm using Libinput with X, so wouldn't X be my compositor?
10:08 vyivel: …ah
10:08 vyivel: hm, idk then
10:09 vyivel: check with xev if the button is correct
10:09 __monty__: In xev I only get NotifyGrab and NotifyUngrab events, no button2 click.
10:10 vyivel: i suppose this means that openbox overrides the middle press action
10:10 vyivel: because it should be just Button{Press,Release} with button 2
10:10 __monty__: It does have bindings for middle press that do work, but its default binding for middle click doesn't.
10:12 vyivel: i'm afraid i'm not familiar with how openbox mouse bindings work
10:13 vyivel: that being said, doesn't seem to be a libinput-related issue
10:13 vyivel: try asking in a some openbox space
10:21 __monty__: OK, thanks for helping to narrow it down.
10:31 whot: __monty__: in X the xf86-input-libinput driver has the libinput context
10:32 whot: the rest still applies, think of libinput as a fancy grep on the device node, starting another grep (like libinput debug-events) is not going to affect the already running one
10:33 whot: from libinput's POV there's no configuration other than what's set at runtime and that's handled by whatever has the context - in the xorg driver case it translates from xorg.conf into the libinput API calls
10:33 whot: debug-events only translates getopt to api calls
12:27 wlb: weston Issue #983 opened by () Is there documentation on how to debug weston? https://gitlab.freedesktop.org/wayland/weston/-/issues/983
12:29 __monty__: whot: And in the absence of /etc/xorg.conf the next likely thing to affect events is the window manager?
18:28 wlb: weston Issue #983 closed \o/ (Asking for help trying to debug MS's weston repo https://gitlab.freedesktop.org/wayland/weston/-/issues/983)
18:30 karenw: Oh yeah, MS has a wayland server. I keep forgetting that in WSL because the wayland socket is in some random directory no app can find.
18:37 bluetail: you make it sound like a dare
23:20 whot: __monty__: if you're on X then typically the desktop environment changes the settings of the xf86-input-* driver, so the actual settings are in e.g. gsettings for gnome
23:21 whot: but the short answer is: if you get NotifyGrab that means something has a grab on the button so it effectively gets intercepted