11:14 zubzub: ifreund: gtk4 and SDL have this weird behaviour
11:16 ifreund: hmm, I'm probably missing something then. jadahl might know why gtk4 does that
13:20 qyliss: Why does libweston have functions like weston_surface_get_content_size() and weston_surface_is_mapped() when surface->width, surface->height, and surface->is_mapped are also in the public headers? Should I be using one over the other?
13:22 daniels: qyliss: please do use the helpers
13:25 qyliss: Is it okay to use struct members when there's no corresponding helper? e.g. I'd like to check whether a surface has been added, so I was thinking of checking whether surface->output was non-null.
14:00 daniels: qyliss: yeah that’s fine if there isn’t a helper, but we should add one
14:01 qyliss: cool, thanks
14:56 wlb: wayland Issue #367 opened by Leah filler-name (sometime) xwayland on nvidia causes flickering and wierd framebuffer bugs https://gitlab.freedesktop.org/wayland/wayland/-/issues/367
16:51 kchibisov: I'd assume compositor must not send negative configure values for xdg_toplevel?
16:52 emersion: yea
16:52 kchibisov: Got [3923471.427] xdg_toplevel@30.configure(-4, -27, array[16]) from sway.
16:53 emersion: sounds like a bug
16:54 kchibisov: I'll redirect the user then.
16:59 jadahl: zubzub: I suspect what you're seeing is eglSwapInterval(0)'s wl_display_sync() in mesa
16:59 kchibisov: What should the client do when gettting such a thing? Should we still ack the configure?
16:59 kchibisov: What we did is acked and then clamped the sizes to min size we request.
17:49 zubzub: jadahl: ah yeah that's probably it. (Why does it do that?)
17:54 jadahl: zubzub: according to the comment it's a way to do at least some throttling when not using the frame callback
17:55 zubzub: o_O