2025/11/25

Newest at the top

2025-11-25 23:13:22 +0100 <hseg> and this is definitively where I give up
2025-11-25 23:12:59 +0100 <hseg> welp. Got so far as https://0x0.st/KWE5.txt
2025-11-25 23:11:09 +0100 <tomsmeding> I meant clear-package-db in the environment file
2025-11-25 23:10:39 +0100merijn(~merijn@host-vr.cgnat-g.v4.dfn.nl) merijn
2025-11-25 23:10:33 +0100 <tomsmeding> that
2025-11-25 23:10:25 +0100 <hseg> ok, by dropping the --package-db=clear parameter from my fix, configure goes through
2025-11-25 23:10:08 +0100 <tomsmeding> isn't the ghcup package db really the package db of the installed GHC, and thus added by default by ghc unless you add this initial command to forget everything (forget what it's called)?
2025-11-25 23:08:35 +0100 <hseg> hrm. the GHC_ENVIRONMENT that cabal passes is incomplete -- it's missing the ghcup package db, which is causing eg base to be reported as missing
2025-11-25 23:06:17 +0100tromp(~textual@2001:1c00:3487:1b00:bd07:71f2:e20b:fdc) (Quit: My iMac has gone to sleep. ZZZzzz…)
2025-11-25 23:06:05 +0100europhobic(~europhobi@109.166.129.14)
2025-11-25 23:04:53 +0100 <hseg> nm re being able to reuse ghc's package-env parser -- Setup.hs configure doesn't use ghc for this
2025-11-25 23:03:28 +0100 <hseg> hrm. OK, shake complains of fewer dependencies missing
2025-11-25 23:02:01 +0100tired-(~tired@user/tired) (Ping timeout: 250 seconds)
2025-11-25 23:00:50 +0100 <tomsmeding> well if it doesn't work without the mess...
2025-11-25 23:00:10 +0100 <hseg> though it's a _mess_
2025-11-25 22:59:58 +0100 <hseg> I should
2025-11-25 22:59:42 +0100 <tomsmeding> hseg: will you be submitting your work back as a PR? :P
2025-11-25 22:59:35 +0100merijn(~merijn@host-vr.cgnat-g.v4.dfn.nl) (Ping timeout: 240 seconds)
2025-11-25 22:56:28 +0100 <hseg> the MaybeT code is still important, though
2025-11-25 22:55:57 +0100 <hseg> after writing a bad parser for the environment files, realized I could just pass them as --ghc-option=-package-env --ghc-option=$GHC_ENVIRONMENT
2025-11-25 22:55:11 +0100merijn(~merijn@host-vr.cgnat-g.v4.dfn.nl) merijn
2025-11-25 22:54:36 +0100hsegfacepalms
2025-11-25 22:52:08 +0100 <hseg> btw, it seems that only by running cabal exec do I get GHC_ENVIRONMENT populated
2025-11-25 22:47:58 +0100 <hseg> hrmph. I broke something
2025-11-25 22:47:12 +0100 <hseg> now to see if my patch actually fixed things...
2025-11-25 22:47:01 +0100 <hseg> well, this bit
2025-11-25 22:46:55 +0100 <hseg> it builds !
2025-11-25 22:46:05 +0100michalz(~michalz@185.246.207.221) (Remote host closed the connection)
2025-11-25 22:46:02 +0100Wygulmage(~Wygulmage@user/Wygulmage) (Quit: Client closed)
2025-11-25 22:43:55 +0100merijn(~merijn@host-vr.cgnat-g.v4.dfn.nl) (Ping timeout: 240 seconds)
2025-11-25 22:42:52 +0100econo_(uid147250@id-147250.tinside.irccloud.com)
2025-11-25 22:42:31 +0100CiaoSen(~Jura@2a02:8071:64e1:da0:5a47:caff:fe78:33db) (Ping timeout: 250 seconds)
2025-11-25 22:39:48 +0100merijn(~merijn@host-vr.cgnat-g.v4.dfn.nl) merijn
2025-11-25 22:36:38 +0100Googulator33(~Googulato@2a01-036d-0106-4ad8-11e2-562e-881b-eb37.pool6.digikabel.hu)
2025-11-25 22:36:23 +0100Googulator33(~Googulato@2a01-036d-0106-4ad8-11e2-562e-881b-eb37.pool6.digikabel.hu) (Quit: Client closed)
2025-11-25 22:36:09 +0100 <tomsmeding> yes :)
2025-11-25 22:33:38 +0100Frostillicus(~Frostilli@71.174.119.69)
2025-11-25 22:29:55 +0100Frostillicus(~Frostilli@pool-71-174-119-69.bstnma.fios.verizon.net) (Ping timeout: 240 seconds)
2025-11-25 22:28:31 +0100merijn(~merijn@host-vr.cgnat-g.v4.dfn.nl) (Ping timeout: 246 seconds)
2025-11-25 22:25:04 +0100 <hseg> ... ah, that's why you wrote fmap .. $ and not <$> -- precedence
2025-11-25 22:21:52 +0100 <hseg> (probably because there are some differences between what I pasted and what I wrote)
2025-11-25 22:21:46 +0100merijn(~merijn@host-vr.cgnat-g.v4.dfn.nl) merijn
2025-11-25 22:21:29 +0100 <hseg> ... for some reason it's not working when I paste it into the actual code -- ghc doesn't think this typechecks?
2025-11-25 22:21:09 +0100 <tomsmeding> I very rarely use MaybeT so I also had to double-check that it does the right thing
2025-11-25 22:11:19 +0100merijn(~merijn@host-vr.cgnat-g.v4.dfn.nl) (Ping timeout: 240 seconds)
2025-11-25 22:10:52 +0100target_i(~target_i@user/target-i/x-6023099) (Quit: leaving)
2025-11-25 22:10:34 +0100 <hseg> it's been a while, am a little rusty
2025-11-25 22:10:31 +0100 <tomsmeding> (because it chooses based on the Maybe, not based on monadic state in m)
2025-11-25 22:10:19 +0100 <hseg> ah, excellent
2025-11-25 22:10:01 +0100 <tomsmeding> but 'MaybeT m' is Alternative regardless of m, as long as m implements Monad