Logs: liberachat/#xmonad
| 2022-02-16 18:58:44 | × | jludwig quits (~justin@user/jludwig) (Quit: ZNC - https://znc.in) |
| 2022-02-16 18:59:37 | → | jludwig joins (~justin@user/jludwig) |
| 2022-02-16 19:29:29 | → | c209e6dc-4d76-47 joins (~aditya@2601:249:4300:1296:195:dac6:592c:a55a) |
| 2022-02-16 19:30:07 | × | c209e6dc-4d76-47 quits (~aditya@2601:249:4300:1296:195:dac6:592c:a55a) (Client Quit) |
| 2022-02-16 19:30:42 | → | c209e6dc-4d76-47 joins (~aditya@2601:249:4300:1296:195:dac6:592c:a55a) |
| 2022-02-16 19:32:43 | × | c209e6dc-4d76-47 quits (~aditya@2601:249:4300:1296:195:dac6:592c:a55a) (Client Quit) |
| 2022-02-16 19:33:19 | → | c209e6dc-4d76-47 joins (~aditya@2601:249:4300:1296:195:dac6:592c:a55a) |
| 2022-02-16 19:33:34 | × | c209e6dc-4d76-47 quits (~aditya@2601:249:4300:1296:195:dac6:592c:a55a) (Client Quit) |
| 2022-02-16 19:34:09 | → | c209e6dc-4d76-47 joins (~aditya@2601:249:4300:1296:195:dac6:592c:a55a) |
| 2022-02-16 19:59:24 | × | Forkk quits (~forkk@li926-228.members.linode.com) (Quit: http://quassel-irc.org - Chat comfortably. Anywhere.) |
| 2022-02-16 20:00:06 | <mc47> | :D why not |
| 2022-02-16 20:02:46 | → | Forkk joins (~forkk@li926-228.members.linode.com) |
| 2022-02-16 20:06:14 | <Solid> | liskin: hah |
| 2022-02-16 20:06:50 | <Solid> | sure, if you're ever in the area let me know and we can go for a coffee :) |
| 2022-02-16 20:10:21 | <Solid> | (I think we can totally spend the money on other "things", like us, because that's why people give us money after all :) |
| 2022-02-16 20:19:28 | <liskin> | yeah, it's just that when we split it between us it's not really that much, so it's more of a "thank you for doing what you'd do anyway" rather than an enablement of some progress |
| 2022-02-16 20:20:44 | <liskin> | that being said, I was going to-sooner or later—suggest that we pay ourselves a bit from that budget as a thank us for 0.17 and 0.17.1 |
| 2022-02-16 20:22:00 | <geekosaur> | I feel like that might be pushing it. paying someone in order to have them put a bit more time into xmonad might be a thing, though |
| 2022-02-16 20:22:16 | <Solid> | this might be the poor student in me speaking, but since we're getting like $850 a month even if we divide that up by a few people that's still pretty decent pocket money |
| 2022-02-16 20:35:38 | × | mc47 quits (~mc47@xmonad/TheMC47) (Remote host closed the connection) |
| 2022-02-16 20:39:24 | <liskin> | yeah, but are you going to spend more time every month working on xmonad? |
| 2022-02-16 20:39:32 | <liskin> | (if yes, then by all means take my share of it :-)) |
| 2022-02-16 20:40:40 | <Solid> | mh, good point |
| 2022-02-16 20:42:48 | <liskin> | hm, is this a good sign for haskell in debian: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1002296 ? |
| 2022-02-16 20:43:11 | <geekosaur> | money wouldhelp me with some things but time working on xmonad wouldn't be one of them, sadly |
| 2022-02-16 20:43:53 | <liskin> | geekosaur: well you already spend a huge amount of time being around on irc helping people (not just in xmonad) :-) |
| 2022-02-16 20:44:17 | <liskin> | (dunno what I wanted to say by that, hm) |
| 2022-02-16 20:44:34 | <liskin> | (probably just kind of supports my previous points) |
| 2022-02-16 20:56:31 | <geekosaur> | mrr. do I want to use cabal properly or lowball it like the stack support? cabal doesn't like working that way, would need cabal exec. |
| 2022-02-16 20:56:46 | geekosaur | is finally getting around to adding cabal support mirroring the stack support |
| 2022-02-16 20:56:57 | <geekosaur> | there being another actual cabal user around finally :) |
| 2022-02-16 21:02:22 | <geekosaur> | hm,that may not work. cabal takes a project file argument but wants a bare filename and searches for it itself |
| 2022-02-16 21:04:02 | <geekosaur> | hah! docs lie. worked fine for list-bin at least |
| 2022-02-16 21:18:08 | <geekosaur> | ok, I have a dumb question. for how long are we going to continue to use ghc --make? |
| 2022-02-16 21:18:34 | <geekosaur> | (--make became the default in ghc 7.0) |
| 2022-02-16 21:24:10 | <geekosaur> | also a not so stupid question. I'm loosely basing my cabal setup on the existing stack one, but if I take my own cabal.project as an example then it would end up building my config twice. Does the typical xmonad stack.yaml contain only the dependencies? |
| 2022-02-16 21:33:08 | <liskin> | we've dropped support for GHCs before 8.4 so we can drop --make right now I think |
| 2022-02-16 21:34:04 | <liskin> | re typical stack.yaml: the setup recommended by INSTALL.md/TUTORIAL.md has deps in stack.yaml and then xmonad.hs is built via stack ghc --make |
| 2022-02-16 21:35:13 | <liskin> | the setup where the user configuration is a something.cabal project itself isn't really documented anywhere (or maybe it's hinted at in xmonad-testing? dunno) |
| 2022-02-16 21:36:22 | <liskin> | not documented, and also not meant to be supported by the ghc/nix/stack recompilation machinery -- a build script is meant to be used for it |
| 2022-02-16 21:37:22 | <liskin> | (or one may skip the recompilation machinery entirely by invoking the xmonad-arch-os binary directly, and then build it whichever way one likes) |
| 2022-02-16 21:43:18 | <geekosaur> | it'sfrom xmonad-testing's example stack-based build script, yes |
| 2022-02-16 21:44:02 | <geekosaur> | not that this isn'tdoable, but I'd been using the xmonad-testing stuff as an example and the stack support code caught me by surprise asa result |
| 2022-02-16 21:50:58 | <liskin> | the xmonad-testing stuff uses a build script, which takes precedence over all that stack support code |
| 2022-02-16 21:51:07 | <liskin> | so not sure why that's an issue |
| 2022-02-16 21:51:28 | <geekosaur> | just that I was using it as a template for how things worked |
| 2022-02-16 22:07:27 | → | ml|_ joins (~ml|@user/ml/x-5298235) |
| 2022-02-16 22:11:01 | × | ml| quits (~ml|@user/ml/x-5298235) (Ping timeout: 272 seconds) |
| 2022-02-16 22:13:33 | <geekosaur> | the --make thing is only relevant because I need to add an option after it if I want to support hvr-style cabal builds at some point |
| 2022-02-16 22:14:14 | <davean> | what are hvr style cabal builds? |
| 2022-02-16 22:15:06 | <davean> | liskin: Its certainly documented - I set mine up that way based off docs in the xmonad repo most of a decade ago - not sure if they're still there or not |
| 2022-02-16 22:16:07 | <geekosaur> | https://github.com/xmonad/xmonad/issues/199 |
| 2022-02-16 22:16:45 | <geekosaur> | basically using a "private" package-env file so we can run ghc directly |
| 2022-02-16 22:17:18 | <geekosaur> | straightforward to add except I think -package-env has to come after --make |
| 2022-02-16 22:19:17 | <liskin> | davean: build script support went in in 2016; I guess it may feel like a decade, but mentioning in anyway just to be sure we're talking about the same thing |
| 2022-02-16 22:19:20 | <geekosaur> | pushed my changes to https://github.com/geekosaur/xmonad/blob/cabal-build/src/XMonad/Core.hs, will test locally before PRing |
| 2022-02-16 22:26:54 | <davean> | liskin: well, I last commited to that git repo in 2015 so. |
| 2022-02-16 22:27:10 | <liskin> | davean: so we're not, I guess? |
| 2022-02-16 22:27:15 | <davean> | I'd guess not |
| 2022-02-16 22:28:34 | <geekosaur> | hm, right, difficult to test when my existing xmonad doesn';t have it yet :) |
| 2022-02-16 22:38:34 | <liskin> | davean: weird; can't find the documentation anywhere in git history :-( |
| 2022-02-16 22:47:20 | <davean> | It was a REALLY long time so, so I'm not sure where I'd find it now |
| 2022-02-16 23:02:31 | <geekosaur> | hm. that's a downside. I get all the xmonad-related stuff but it can't find dbus without a cabal file |
| 2022-02-16 23:03:25 | <geekosaur> | may mean I just can't get away with that support although most folks can, or this is just a difference between cabal.project and stack.yaml that's not easily worked around |
| 2022-02-16 23:04:12 | × | werneta quits (~werneta@70-142-214-115.lightspeed.irvnca.sbcglobal.net) (Ping timeout: 272 seconds) |
| 2022-02-16 23:05:20 | → | werneta joins (~werneta@70-142-214-115.lightspeed.irvnca.sbcglobal.net) |
| 2022-02-16 23:08:57 | <liskin> | not sure what cabal can't and can't do, but with stack, you can do stack build totally-unrelated-package and that package is then visible in the subsequent stack (exec) ghc |
| 2022-02-16 23:09:32 | <liskin> | and I think this isn't really intended, it's more of a side-effect of the way stack uses ghc pkg envs |
| 2022-02-16 23:23:13 | <geekosaur> | okay, PRed |
| 2022-02-16 23:23:34 | <geekosaur> | someone with a cabal setup and no extra deps will have to test it though |
| 2022-02-16 23:26:46 | <geekosaur> | hvr-style could deal with this though, I'd "just" cabal install --lib --package-env=$HOME/.config/xmonad dbus dbus-client |
| 2022-02-16 23:29:33 | <liskin> | so what are the pros/cons of this new things compared to --package-env? |
| 2022-02-16 23:35:32 | × | jmct_ quits (sid160793@id-160793.tinside.irccloud.com) (Ping timeout: 240 seconds) |
| 2022-02-16 23:35:48 | → | jmct_ joins (sid160793@id-160793.tinside.irccloud.com) |
| 2022-02-16 23:35:49 | × | PotatoGim quits (sid99505@id-99505.lymington.irccloud.com) (Ping timeout: 240 seconds) |
| 2022-02-16 23:36:03 | × | pl quits (sid98063@id-98063.helmsley.irccloud.com) (Read error: Connection reset by peer) |
| 2022-02-16 23:36:41 | → | pl joins (sid98063@id-98063.helmsley.irccloud.com) |
| 2022-02-16 23:36:55 | → | PotatoGim joins (sid99505@id-99505.lymington.irccloud.com) |
| 2022-02-16 23:39:38 | × | seschwar quits (~seschwar@user/seschwar) (Quit: :wq) |
| 2022-02-16 23:41:14 | <geekosaur> | it works just like stack.yaml support, basically. which amoing other things means no manual "cabal install --lib --package-env=... xmonad xmonad-contrib foo bar ..." needed |
| 2022-02-16 23:41:39 | <geekosaur> | which makes it less flexible but easier to set up |
| 2022-02-16 23:42:08 | <geekosaur> | hm, I guess one downside is you need a local checkout of xmonad and xmonad-contrib |
| 2022-02-16 23:42:44 | <geekosaur> | since unlike stack.yaml there is no list of packages aside from the locally built onesin cabal.project |
| 2022-02-16 23:43:03 | <geekosaur> | which is the same place my dbus issue came from |
| 2022-02-16 23:43:23 | <geekosaur> | that said, I don't think you really want a cabal setup unless you *are* doing that |
| 2022-02-16 23:43:36 | <geekosaur> | and probably same for stack |
| 2022-02-16 23:44:21 | <geekosaur> | I'll add support for the other one later, this one was basically to warm up |
| 2022-02-16 23:54:12 | <liskin> | what "other one" ? |
| 2022-02-16 23:54:30 | <geekosaur> | the package-env version |
| 2022-02-16 23:54:47 | <liskin> | hm, does that one need any more than what we already have? |
| 2022-02-16 23:55:05 | <geekosaur> | basically test for existence of the package-env file and arrange to append -package-env=... to the ghc build |
| 2022-02-16 23:55:23 | <liskin> | I thought ghc picks that up automatically |
| 2022-02-16 23:55:45 | <geekosaur> | we're using a custom package env, specifically so it doesn't mess with other builds |
| 2022-02-16 23:55:47 | <liskin> | otherwise the steps in INSTALL.md wouldn't work |
| 2022-02-16 23:57:21 | <geekosaur> | hm, maybe it will then |
| 2022-02-16 23:57:43 | <geekosaur> | that would be good, just have to document needfor additional commands for cases like mine |
| 2022-02-16 23:58:20 | <liskin> | (in unrelated news, the xmobar memleak is actually in xmobar, I just spent an entire day trying to write a small C reproducer, and after failing and doing some gdb digging in a leaking xmobar, concluded it's actually xmobar's fault after all :-)) |
| 2022-02-16 23:58:21 | <geekosaur> | hvr's stuff implied we needed to specify it |
| 2022-02-16 23:59:08 | <liskin> | well hvr's last comment explained that specifying --package-env=~/.xmonad will make things just work |
| 2022-02-16 23:59:21 | <liskin> | and after I've verified that I turned that into instructions in INSTALL.md |
All times are in UTC.