Logs: liberachat/#xmonad
| 2022-02-28 00:14:42 | <geekosaur> | mod-q should be sufficient |
| 2022-02-28 00:15:27 | <iqubic> | Cool. |
| 2022-02-28 00:16:46 | <iqubic> | I've figured out the issue here... |
| 2022-02-28 00:17:09 | <iqubic> | I have "isFullscreen --> doFullFloat" as one of my manageHook rules. |
| 2022-02-28 00:18:36 | <geekosaur> | that actually shouldnot be involved here since I don't think firefox withdraws its window when it fullscreens |
| 2022-02-28 00:19:00 | <geekosaur> | and in any case doFullFloat ignores struts |
| 2022-02-28 00:23:59 | <iqubic> | If have two firefox windows open side by side, and I tell the left one to become fullscreen, it only uses its allotted rectangle of space to go fullscreen, as can be seen here: https://imgur.com/a/DfBM2oY |
| 2022-02-28 00:24:09 | <iqubic> | So, it's not floating at all. |
| 2022-02-28 00:25:01 | <iqubic> | This is why I don't think my manageHook rule of "isFullscreen --> doFullFloat" is working. |
| 2022-02-28 00:26:11 | <geekosaur> | as I just said, manageHooks don't fire in that case. Firefox would have to withdraw the window and remap it with the STATE_FULLSCREEN atom attached |
| 2022-02-28 00:26:33 | <iqubic> | Oh. I see. That's an issue. |
| 2022-02-28 00:26:36 | <geekosaur> | what Firefox is doing is handled by an event instead |
| 2022-02-28 00:26:45 | <iqubic> | Is there anything I can do to fix this? |
| 2022-02-28 00:26:46 | <geekosaur> | and that event is what ewmhFullscreen watches for |
| 2022-02-28 00:27:28 | <geekosaur> | that said, the default action should be to fullscreen it without regard for struts, like it does here, iirc |
| 2022-02-28 00:27:46 | <iqubic> | My main looks like this. https://dpaste.com/H5EXBTRVV |
| 2022-02-28 00:27:58 | <iqubic> | I'm very clearly using ewmhFullscreen. |
| 2022-02-28 00:28:59 | <iqubic> | But I'm still getting the behavior shown in my latest screenshot. |
| 2022-02-28 00:32:14 | <iqubic> | geekosaur: do you know what's going on here? |
| 2022-02-28 00:44:16 | <geekosaur> | hm. you have ewmhFullscreen but not ewmh |
| 2022-02-28 00:44:40 | <geekosaur> | docks $ ewmh $ ewmhFullscreen def |
| 2022-02-28 00:45:29 | <iqubic> | Yeah, I just added ewmh too, and I'm still getting the same behavior. |
| 2022-02-28 00:46:22 | <geekosaur> | I'm lost then |
| 2022-02-28 00:46:35 | <iqubic> | Yeah... this is weird. |
| 2022-02-28 00:47:09 | → | triumphofdeath[m joins (~triumphof@2001:470:69fc:105::a5c) |
| 2022-02-28 00:50:14 | <geekosaur> | mm, I odn'thave ewmh in mine either, but my config is based on desktopConfig which already includes it |
| 2022-02-28 00:50:55 | <geekosaur> | I doubt making that change would help you though |
| 2022-02-28 00:54:29 | <iqubic> | No. That didn't help. |
| 2022-02-28 00:55:16 | <iqubic> | This is a strange issue.. |
| 2022-02-28 01:00:57 | <iqubic> | This is an odd issue. Why is firefox being a nuisance here? |
| 2022-02-28 01:01:57 | <geekosaur> | I have no idea |
| 2022-02-28 01:02:36 | <geekosaur> | hm. I guess the other key difference here is I run chrome instead of firefox. can try launching it in a spare workspace and see if it behaves |
| 2022-02-28 01:03:04 | <iqubic> | That might help. |
| 2022-02-28 01:03:15 | <geekosaur> | that works here too |
| 2022-02-28 01:03:37 | <iqubic> | Weird... Is your config public? Can I see? |
| 2022-02-28 01:04:48 | <geekosaur> | https://github.com/geekosaur/xmonad.hs/blob/skkukuk/xmonad.hs if you dare:) |
| 2022-02-28 01:06:56 | → | mvk joins (~mvk@2607:fea8:5cc3:7e00::8c21) |
| 2022-02-28 01:08:03 | <iqubic> | is mateConfig different from desktopConfig in any substantial ways? |
| 2022-02-28 01:08:39 | <geekosaur> | it uses mate run and logout dialogs instead of the standard ones, otherwise no |
| 2022-02-28 01:08:45 | <iqubic> | I see. |
| 2022-02-28 01:10:12 | <iqubic> | I notice that "isFullscreen --> doFullFloat" isn't in your config. Why is that? |
| 2022-02-28 01:10:57 | <geekosaur> | I don't use anything that maps itself fullscreened |
| 2022-02-28 01:11:05 | <geekosaur> | if I did I'dprobably add it |
| 2022-02-28 01:11:53 | <iqubic> | But you do use fullscreen applications, right? Like hitting F11 in Chrome makes that fullscreen for you, right? |
| 2022-02-28 01:12:37 | <geekosaur> | yes. also in my IRC client |
| 2022-02-28 01:12:54 | <iqubic> | What part of your config is controlling that? |
| 2022-02-28 01:13:12 | <geekosaur> | as I mentioned earlier, both of those use events (ewmhFullscreen) instead of unmapping and remapping (which is the only time the manageHook runs) |
| 2022-02-28 01:13:39 | <iqubic> | Firefox also uses events, right? |
| 2022-02-28 01:14:03 | <iqubic> | So in theory, this should just work... |
| 2022-02-28 01:14:17 | <geekosaur> | right. and it does work here |
| 2022-02-28 01:14:32 | <geekosaur> | with chrome, with firefox, with hexchat |
| 2022-02-28 01:14:32 | <iqubic> | What's different on my system? |
| 2022-02-28 01:14:53 | <iqubic> | I'm so confused. |
| 2022-02-28 01:16:03 | <iqubic> | Chromium uses event right? |
| 2022-02-28 01:16:34 | <geekosaur> | all three of the programs I mentioned use events |
| 2022-02-28 01:16:34 | <iqubic> | I just installed Chromium on my machine, to test... That goes fullscreen just fine! |
| 2022-02-28 01:16:43 | <geekosaur> | weird |
| 2022-02-28 01:17:04 | <geekosaur> | although, you had been missing ewmh before. what happens if you restart firefox? |
| 2022-02-28 01:18:01 | <iqubic> | I'll be back in a moment. I haven't set up an IRC client on this machine, so I'm using a web based client in firefox. BRB. |
| 2022-02-28 01:18:06 | × | iqubic quits (~iqubic@2601:602:9502:c70::7507) (Quit: Client closed) |
| 2022-02-28 01:18:33 | → | iqubic joins (~iqubic@2601:602:9502:c70::7507) |
| 2022-02-28 01:18:36 | <geekosaur> | I suspect firefox only checks _NET_SUPPORTED on startup, and you've made several changes to it so far by upgrading xmonad and changing ewmh and ewmhFullscreen |
| 2022-02-28 01:19:14 | <iqubic> | Well, I just restarted firefox. Still not working. :( |
| 2022-02-28 01:19:57 | <iqubic> | This is very very weird.... |
| 2022-02-28 01:23:31 | <iqubic> | I'm so very confused. |
| 2022-02-28 01:29:08 | <iqubic> | xmonad is a lovely window manager. There's just this one issue. |
| 2022-02-28 01:30:42 | <iqubic> | I'm so very flummoxed here. |
| 2022-02-28 01:38:27 | <iqubic> | geekosaur: Just on a whim, I tried changing the value of startupHook from "startupHook = myStartupHook" to "myStartupHook <+> startupHook defConfig". Recall that "defConfig = dock $ ewmh $ ewmhFullscreen" Now firefox is able to go fullscreen just fine. |
| 2022-02-28 01:39:08 | <iqubic> | Why the heck would that make any difference, and why would that only affect firefox and not chromium? |
| 2022-02-28 01:42:02 | <fizzie> | `ewmhFullscreen` at least adds a startupHook that does the EWMH fullscreen advertisement. |
| 2022-02-28 01:42:55 | <iqubic> | Make sense. |
| 2022-02-28 01:43:00 | <fizzie> | (As in, adds the _NET_WM_STATE and _NET_WM_STATE_FULLSCREEN atoms to the _NET_SUPPORTED property of the root window.) |
| 2022-02-28 01:44:14 | <geekosaur> | but that is added *to* the startupHook, you shouldn't need that |
| 2022-02-28 01:44:37 | <geekosaur> | oh wait, but startupHook defConfig. yes, that makes sense |
| 2022-02-28 01:46:03 | <iqubic> | Yeah. I'm now mconcat-ing my own startup hook to the startup hooks generated via (docks $ ewmh $ ewmhFullscreen def) |
| 2022-02-28 01:47:12 | <iqubic> | <+> is just infix mconcat, right? Why does that exist if <> is also infix mconcat and already in Data.Monoid? Is this just a re-export with a new name so that new users don't need to import Data.Monoid? |
| 2022-02-28 01:47:49 | <geekosaur> | <+> came first, and <> was actually borrowed from xmonad |
| 2022-02-28 01:48:06 | <iqubic> | Wait... really? That's so wild. |
| 2022-02-28 01:48:09 | <geekosaur> | they renamed it because ghc was already using <+> with a different meaning (from one of the prettyprinter packages iirc) |
| 2022-02-28 01:51:24 | × | jao quits (~jao@cpc103048-sgyl39-2-0-cust502.18-2.cable.virginm.net) (Ping timeout: 240 seconds) |
| 2022-02-28 01:55:24 | <geekosaur> | in the first few releases of xmonad, <+> was mconcat specialized to ManageHooks. That changed around 2008 or 2009 when we removed the specialization and just made it mconcat. <> was around 2013, I think |
| 2022-02-28 02:02:43 | <geekosaur> | also it's mappend, not mconcat,whoops. composeAll is mconcat specialized to ManageHooks |
| 2022-02-28 02:07:17 | → | catman joins (~catman@user/catman) |
| 2022-02-28 02:15:04 | <iqubic> | geekosaur: Yeah, I meant mconcat. |
| 2022-02-28 02:18:09 | <catman> | got this error on both arch and artix after a haskell update with pacman. not entirely sure how to resolve. |
| 2022-02-28 02:18:16 | <catman> | xmonad: error while loading shared libraries: libHSutf8-string-1.0.2-9eFA57BgKXNHtjvfuIOq9f-ghc9.0.2.so: cannot open shared object file: No such file or directory |
| 2022-02-28 02:19:00 | <catman> | have tried reinstalling xmonad-contrib and ghc |
| 2022-02-28 02:23:04 | × | catman quits (~catman@user/catman) (Quit: WeeChat 3.5-dev) |
| 2022-02-28 02:55:46 | × | iqubic quits (~iqubic@2601:602:9502:c70::7507) (Quit: Client closed) |
| 2022-02-28 02:59:07 | <geekosaur> | @tell catman arch screwed something up again. if they handled their haskell packages properly this wouldn't happen. (you're the second one tonight) |
| 2022-02-28 02:59:07 | <lambdabot> | Consider it noted. |
| 2022-02-28 03:00:31 | <geekosaur> | @tell catman if you want reliably working xmonad and xmobar, wipe all haskell related arch packages from your system, use ghcup to install ghc and stack and/or cabal, then follow https://xmonad.org/INSTALL.html |
| 2022-02-28 03:00:31 | <lambdabot> | Consider it noted. |
| 2022-02-28 03:03:00 | × | steve__ quits (~steve@ool-182c2b80.dyn.optonline.net) (Ping timeout: 240 seconds) |
| 2022-02-28 03:03:17 | × | banc quits (banc@gateway/vpn/airvpn/banc) (Ping timeout: 272 seconds) |
| 2022-02-28 03:21:01 | × | td_ quits (~td@muedsl-82-207-238-210.citykom.de) (Ping timeout: 272 seconds) |
| 2022-02-28 03:22:33 | → | td_ joins (~td@94.134.91.216) |
| 2022-02-28 03:25:13 | → | banc joins (banc@gateway/vpn/airvpn/banc) |
| 2022-02-28 03:58:42 | → | iqubic joins (~iqubic@c-67-171-37-233.hsd1.wa.comcast.net) |
| 2022-02-28 04:00:44 | <iqubic> | Is it possible to override the borderWidth value for certain windows? Specifically I'd like to remove the border from Firefox's right click context menu (why is this even a separate window?), and also from the windows spawned by the dunst notification daemon. |
All times are in UTC.