Logs: liberachat/#xmonad
| 2021-10-08 16:40:25 | → | werneta joins (~werneta@70-142-214-115.lightspeed.irvnca.sbcglobal.net) |
| 2021-10-08 16:53:37 | <FOSSHuman[m]> | I am definitely going to sponsor XMonad when I have some money, XMonad-contrib is really amazing aswell |
| 2021-10-08 17:00:36 | × | werneta quits (~werneta@70-142-214-115.lightspeed.irvnca.sbcglobal.net) (Remote host closed the connection) |
| 2021-10-08 17:01:45 | <liskin> | https://github.com/liskin/dotfiles/commit/75dfc057c33480ee9d3300d4d02fb79a986ef3a5#diff-87ceeecfb90f21378eea9decafa779fe68bd659b3b8802982639baa30c504b8f |
| 2021-10-08 17:01:51 | <liskin> | what have I done :-D |
| 2021-10-08 17:03:01 | <liskin> | (I could probably write a blog post describing how to write xmobar plugins in bash, but… time) |
| 2021-10-08 17:04:00 | <mc47> | God :D |
| 2021-10-08 17:04:08 | <mc47> | that's brilliant |
| 2021-10-08 17:06:10 | → | werneta joins (~werneta@70-142-214-115.lightspeed.irvnca.sbcglobal.net) |
| 2021-10-08 17:06:19 | <mc47> | I'm not a bash wizard but I can kinda see what you're doing |
| 2021-10-08 17:13:41 | geekosaur | would probably do it in perl, but. |
| 2021-10-08 17:14:05 | <geekosaur> | (or in haskell, it's fairly simple after all) |
| 2021-10-08 17:18:21 | <jakefromstatefar> | Or nim, it's a nice scripting lang. |
| 2021-10-08 17:23:18 | <Solid> | liskin: that is so horrible it's amazing :D |
| 2021-10-08 17:24:19 | <Solid> | (btw, according to my logs LSLeary was last seen on 2020-03-13 (this seems oddly pandemic related, I hope it's not)) |
| 2021-10-08 17:24:37 | <geekosaur> | I didn't see anything horrible about it (aside from, well, bash) |
| 2021-10-08 17:24:50 | <geekosaur> | clear and concice, as bash code goes |
| 2021-10-08 17:25:02 | <geekosaur> | *concise |
| 2021-10-08 17:25:22 | <Solid> | not horrible as in "horrible code", more like horrible as in "writing an xmobar plugin in bash" :) |
| 2021-10-08 17:27:44 | <jakefromstatefar> | For those not in #waymonad:matrix.org, but still interested: |
| 2021-10-08 17:27:47 | jakefromstatefar | uploaded an image: (179KiB) < https://libera.ems.host/_matrix/media/r0/download/jupiterbroadcasting.com/twFQGzGRStvJPlBwaXkkuQSb/image.png > |
| 2021-10-08 17:30:04 | <geekosaur> | hm, I was under the impression they wanted to start over because wlroots was an ugly broken hack |
| 2021-10-08 17:33:52 | <jakefromstatefar> | ¯\_(ツ)_/¯ |
| 2021-10-08 17:37:13 | <liskin> | I don't really know why, but I consider perl to be a bit of a dead end these days, so I don't use it any more; also, I absolutely hate the way function arguments are passed in it |
| 2021-10-08 17:38:05 | <liskin> | it'd probably be more "right" to write it in haskell, but I'd have to migrate my xmobar config from xmobarrc to compiled, and then… it'd almost certainly be longer |
| 2021-10-08 17:38:28 | <liskin> | bash is fairly concise once one learns it properly |
| 2021-10-08 17:45:36 | <Solid> | that's fair |
| 2021-10-08 17:55:02 | <geekosaur> | I'm an olde phart™ so I still like perl with all its freakishness |
| 2021-10-08 17:55:25 | <geekosaur> | then again I invented the perl 4 hack for nested structures, so. |
| 2021-10-08 17:56:17 | <liskin> | which perl 4 hack? |
| 2021-10-08 17:56:20 | <geekosaur> | glob abuse, mm |
| 2021-10-08 17:57:07 | <geekosaur> | storing the glob reference of one structure inside another. had to do an extra glob deref to use it, but it was cleaner than the other ways of doing it |
| 2021-10-08 17:57:18 | <geekosaur> | and in some sense inspired perl 5 refs |
| 2021-10-08 17:58:11 | <liskin> | hm, I think I'm not old enough to have ever used perl 4, actually |
| 2021-10-08 17:58:26 | <geekosaur> | I started out with perl 3 |
| 2021-10-08 17:58:47 | <geekosaur> | which I knew about because I was moderating comp.sources.misc whne it was released |
| 2021-10-08 17:58:54 | <liskin> | I think I had a perl 4 book when I was like 15 years old or something, but when I started coding in Perl around 2002 or 3, it was probably already perl 5 |
| 2021-10-08 18:00:29 | <geekosaur> | first public release of python was the same week but I couldn't get into it; python 1 was rather limited |
| 2021-10-08 18:00:56 | <geekosaur> | only remotely interesting part of it to me was STDWIN and it didn't take me long to figure out it was a horrid hack |
| 2021-10-08 18:01:05 | <geekosaur> | (long since disowned) |
| 2021-10-08 18:08:47 | × | geekosaur quits (~geekosaur@xmonad/geekosaur) (Remote host closed the connection) |
| 2021-10-08 18:10:37 | → | geekosaur joins (~geekosaur@xmonad/geekosaur) |
| 2021-10-08 18:21:33 | × | mc47 quits (~mc47@xmonad/TheMC47) (Remote host closed the connection) |
| 2021-10-08 18:47:49 | × | wonko quits (~wjc@62.115.229.50) (Ping timeout: 252 seconds) |
| 2021-10-08 18:49:15 | → | wonko joins (~wjc@62.115.229.50) |
| 2021-10-08 19:00:06 | × | werneta quits (~werneta@70-142-214-115.lightspeed.irvnca.sbcglobal.net) (Remote host closed the connection) |
| 2021-10-08 19:06:09 | Hash | is now known as RitaDivaj |
| 2021-10-08 19:07:48 | yuu[m] | uploaded a video: (1330KiB) < https://libera.ems.host/_matrix/media/r0/download/matrix.org/enZFitgUqQxjJPGAsNPbpxXJ/ffmpeg-2021-10-08T18%3A58%3A23%2B00%3A00.mp4 > |
| 2021-10-08 19:07:49 | <yuu[m]> | Given a workspace with 2+ windows A (master) and B in tabbed layout, and 1 floating window F. Current focus is on B. If the user changes the focus to F, xmonad changes the tabbed focus to A. |
| 2021-10-08 19:07:49 | <yuu[m]> | Any idea how to fix/workaround that? |
| 2021-10-08 19:08:03 | <yuu[m]> | s/+// |
| 2021-10-08 19:10:43 | <geekosaur> | TrackFloating module might help, otherwise no, our float handling is atrocious and hard to fix |
| 2021-10-08 19:11:09 | <jakefromstatefar> | Hooks.RefocusLast and Layout.TrackFloating |
| 2021-10-08 19:11:32 | <jakefromstatefar> | geekosaur: Speaking of, how would a tree datatype solve this issue? |
| 2021-10-08 19:11:42 | <jakefromstatefar> | Just give a separate branch for floating windows? |
| 2021-10-08 19:12:16 | <geekosaur> | not sure, it was liskin's idea not mine. in any case tree datatype is for layouts not windows so I'm not sure it affects it |
| 2021-10-08 19:12:40 | <geekosaur> | unless there's a distinct floating layout somrewhere instead of just a map of windows to rects |
| 2021-10-08 19:14:36 | <liskin> | no, the tree would actually hold both layouts and windows |
| 2021-10-08 19:15:13 | <liskin> | so we'd have a root node that wouldn't do nothing special except layout all subnodes with the full rect and then stack the resulting windows on top of one another |
| 2021-10-08 19:15:38 | <jakefromstatefar> | How would it hold layouts? Just as an optional attribute at each branch, holding the layout function? |
| 2021-10-08 19:15:39 | <liskin> | under this node there'd be a floating layout node, and all windows under this node would be in the floating layer, and it'd have its own focus |
| 2021-10-08 19:15:55 | <liskin> | another node under the root would be the "primary" layout, where most windows will go |
| 2021-10-08 19:16:13 | <jakefromstatefar> | OK, how would having 2 focus' work? |
| 2021-10-08 19:16:14 | <liskin> | again, this node would remember its own focus |
| 2021-10-08 19:16:51 | <liskin> | every node would just remember which of its subnodes has focus |
| 2021-10-08 19:17:07 | <liskin> | thus there'd be a single window having focus |
| 2021-10-08 19:17:08 | <jakefromstatefar> | Could you pseudo-yaml format this idea? I'm having a hard time understanding the fullness of this concept. |
| 2021-10-08 19:17:35 | <liskin> | I can try |
| 2021-10-08 19:18:18 | <jakefromstatefar> | liskin: The floating layer, and sublayout layers, both hold a focus, right? |
| 2021-10-08 19:26:45 | <geekosaur> | yes. and presumably only one of these layers has the focus at any one time |
| 2021-10-08 19:27:21 | <geekosaur> | at one point dons tried to implement that with multiple StackSets; it didn't go well. but then the StackSet is arguably a large part of the problem |
| 2021-10-08 19:27:26 | RitaDivaj | is now known as Hash |
| 2021-10-08 19:29:41 | <jakefromstatefar> | Mmh, a thought just came to my head... Could this tree approach allow for shareable multiseat solution? |
| 2021-10-08 19:30:21 | <jakefromstatefar> | Basically, just allowing any node to set the `DISPLAY` value. |
| 2021-10-08 19:31:01 | <jakefromstatefar> | And, then you could have multiple root nodes, each linked to a different TTY, or physical display. |
| 2021-10-08 19:31:07 | <jakefromstatefar> | And, have the ability to interop with eachother. |
| 2021-10-08 19:31:10 | <liskin> | https://on.tty-share.com/s/9uow43Ip7P0KXGKfw3zcU1w0RptHCDrbhw_2HoJhXdrfnnHHSOOYDXKnDzldl4lh264/ |
| 2021-10-08 19:31:22 | <liskin> | (I may need to refresh it once you connect) |
| 2021-10-08 19:31:49 | <geekosaur> | jakefromstatefar, interop would be difficult because each window belongs to a specific display |
| 2021-10-08 19:31:56 | <geekosaur> | enforced by X11 |
| 2021-10-08 19:33:46 | <jakefromstatefar> | Hm, perhaps we could just leave it as an unimplemented path forward, and think more on it if/when we get rolling with wayland. |
| 2021-10-08 19:33:52 | <jakefromstatefar> | Not sure what the policy is there, though. |
| 2021-10-08 19:34:40 | <geekosaur> | I plan to collect the discussions here and create a WIP placeholder issue |
| 2021-10-08 19:34:59 | <geekosaur> | which we can brainstorm on as and when we get time/motivation |
| 2021-10-08 19:35:23 | <liskin> | geekosaur: I'll send you the yaml once I'm finished with it |
| 2021-10-08 19:35:25 | <jakefromstatefar> | liskin: What of allowing decoration as a separate function per-node. Meaning tabbed would be a core feature, and not something requiring more effort. |
| 2021-10-08 19:36:10 | × | zouhair quits (~zouhair@142.182.58.252) (Ping timeout: 252 seconds) |
| 2021-10-08 19:36:31 | <liskin> | jakefromstatefar: yes tabbing would fit naturally into this |
| 2021-10-08 19:36:49 | <liskin> | I just added a sample tabbed panes into my pseudo-yaml |
| 2021-10-08 19:37:05 | <jakefromstatefar> | What if instead of using an int focus value, using a stackset/other holepunched list datatype for the windows? |
| 2021-10-08 19:37:22 | <liskin> | that's an implementation detail mostly |
| 2021-10-08 19:37:57 | <liskin> | for the sizes of structures that xmonad typically deals with, it doesn't really matter which is which |
| 2021-10-08 19:38:08 | → | zouhair joins (~zouhair@142.182.58.252) |
| 2021-10-08 19:39:10 | <jakefromstatefar> | What of background setting? |
| 2021-10-08 19:39:25 | <jakefromstatefar> | Are we going to include a window node for the root window inside the root node? |
| 2021-10-08 19:40:09 | <liskin> | geekosaur: https://x0.at/fh6O.txt is what I have now |
| 2021-10-08 19:40:13 | <yuu[m]> | <geekosaur> "TrackFloating module might help,..." <- i added trackFloating to my layout hook and it was able to handle it. thank you very much! |
| 2021-10-08 19:40:56 | <liskin> | jakefromstatefar: root window is unmanaged in the X world, so no |
| 2021-10-08 19:41:39 | <liskin> | we don't need to concern ourselves with setting the background |
| 2021-10-08 19:42:00 | <jakefromstatefar> | ah |
All times are in UTC.