2026/02/17

2026-02-17 01:15:38 +0100hightower4(~hightower@dh207-83-96.xnet.hr) (Ping timeout: 252 seconds)
2026-02-17 01:19:56 +0100hightower3(~hightower@dh207-83-96.xnet.hr) hightower2
2026-02-17 02:11:48 +0100beastwick(~brian@user/beastwick) beastwick
2026-02-17 02:13:46 +0100 <beastwick> hello, not sure if this is a bug, but I am simply swapping screens with XMonad.Actions.CycleWS swapNextScreen. On one monitor I have tabbed layout, so multiple windows with tabs at the top. When I swap, the workspaces correctly switch, but the tab bar appears over the other workspace and can only be removed if I restart xmonad in place.
2026-02-17 02:14:06 +0100werneta(~werneta@71.83.160.242) werneta
2026-02-17 02:15:53 +0100 <beastwick> oh wait, now it isn't happening not sure why
2026-02-17 02:16:04 +0100 <beastwick> weird
2026-02-17 02:23:11 +0100 <geekosaur> known bug, some layout modifier you're using isn't passing on the Hide message to Tabbed
2026-02-17 02:23:34 +0100 <beastwick> oh :(, can I pass it along, or modify the function?
2026-02-17 02:23:36 +0100 <geekosaur> sadly a lot of layout modifiers don't pass on the Hide or ReleaseResources messages
2026-02-17 02:28:51 +0100 <geekosaur> you'd have to bisect to find the layout modifier that isn't passing them on and make a local copy of it (in ~/.xmonad/lib/XMonad/Layout/whatever.hs) with the handleMessage logic added (see for example https://github.com/xmonad/xmonad-contrib/pull/813/changes#diff-39a8849b1ce55ed5ed86080ca98a422374c…)
2026-02-17 02:29:14 +0100 <beastwick> thanks, this is what I am going to do
2026-02-17 02:29:20 +0100 <beastwick> I already have some other modified classes
2026-02-17 02:29:36 +0100 <geekosaur> if a layout modifier uses the XMonad.Layout.LayoutModifier framework, that already handles them for it
2026-02-17 02:30:33 +0100 <beastwick> oh so, cyclews swapNextScreen can't ensure this, I have to fix the layout modifier itself?
2026-02-17 02:31:00 +0100 <geekosaur> (also if you scroll up on that link, you'll find the old version which actually has the bug in question; note that its handleMess deals only with its own message. I should probably fix that before liskin gets around to reviewing the whole thing)