2025/11/25

Newest at the top

2025-11-25 20:44:33 +0100 <tomsmeding> (build it first with `cabal build`, then run with `cabal exec`)
2025-11-25 20:44:16 +0100 <tomsmeding> and under `cabal exec`?
2025-11-25 20:44:02 +0100 <hseg> nope -- dumping environment under cabal run shows no GHC_* envvars
2025-11-25 20:43:35 +0100pavonia(~user@user/siracusa) siracusa
2025-11-25 20:40:06 +0100Frostillicus(~Frostilli@pool-71-174-119-69.bstnma.fios.verizon.net)
2025-11-25 20:39:35 +0100Frostillicus(~Frostilli@pool-71-174-119-69.bstnma.fios.verizon.net) (Read error: Connection reset by peer)
2025-11-25 20:39:19 +0100merijn(~merijn@host-vr.cgnat-g.v4.dfn.nl) (Ping timeout: 244 seconds)
2025-11-25 20:37:12 +0100 <geekosaur> hack, see if it's there with `cabal run`? (there are enough differences between the two that using `cabal run` to run test suites is relatively common)
2025-11-25 20:35:06 +0100merijn(~merijn@host-vr.cgnat-g.v4.dfn.nl) merijn
2025-11-25 20:28:10 +0100ttybitnik(~ttybitnik@user/wolper) (Quit: Fading out...)
2025-11-25 20:25:54 +0100tomsmedingwonders how long ago it has been that this has worked properly in the first place
2025-11-25 20:24:08 +0100 <hseg> Giving up, I'll use something else then
2025-11-25 20:23:55 +0100merijn(~merijn@host-vr.cgnat-g.v4.dfn.nl) (Ping timeout: 240 seconds)
2025-11-25 20:19:44 +0100merijn(~merijn@host-vr.cgnat-g.v4.dfn.nl) merijn
2025-11-25 20:19:39 +0100mulk(~mulk@pd9514972.dip0.t-ipconnect.de) mulk
2025-11-25 20:18:51 +0100 <hseg> urggh
2025-11-25 20:18:46 +0100 <hseg> hrmph. it seems cabal _builds_ the test suite with GHC_ENVIRONMENT set, but doesn't _run_ it with it
2025-11-25 20:18:01 +0100jmcantrell(~weechat@user/jmcantrell) (Ping timeout: 264 seconds)
2025-11-25 20:16:02 +0100tromp(~textual@2001:1c00:3487:1b00:bd07:71f2:e20b:fdc)
2025-11-25 20:14:04 +0100peterbecich(~Thunderbi@172.222.148.214) (Ping timeout: 256 seconds)
2025-11-25 20:13:02 +0100tomsmedingwishing hseg luck from the sideline
2025-11-25 20:11:34 +0100 <hseg> .... OK, we're gradually closing in on a culprit -- it seems GHC_ENVIRONMENT is getting unset somewhere along the line
2025-11-25 20:11:31 +0100jmcantrell(~weechat@user/jmcantrell) jmcantrell
2025-11-25 20:10:55 +0100merijn(~merijn@host-vr.cgnat-g.v4.dfn.nl) (Ping timeout: 240 seconds)
2025-11-25 20:06:41 +0100merijn(~merijn@host-vr.cgnat-g.v4.dfn.nl) merijn
2025-11-25 20:06:13 +0100kitadamian(~kitadamia@95.49.84.49.ipv4.supernova.orange.pl)
2025-11-25 20:04:36 +0100Frostillicus(~Frostilli@pool-71-174-119-69.bstnma.fios.verizon.net)
2025-11-25 20:04:14 +0100 <hseg> after all, GHC reads it from its environment
2025-11-25 20:04:01 +0100 <hseg> ... wait, shouldn't it Just Work?
2025-11-25 20:03:59 +0100Everything(~Everythin@172-232-54-192.ip.linodeusercontent.com) Everything
2025-11-25 20:03:55 +0100Frostillicus(~Frostilli@pool-71-174-119-69.bstnma.fios.verizon.net) (Ping timeout: 240 seconds)
2025-11-25 20:02:54 +0100Everything(~Everythin@172-232-54-192.ip.linodeusercontent.com) (Quit: leaving)
2025-11-25 20:02:31 +0100 <hseg> hrm. any way of making Setup.hs read GHC_ENVIRONMENT? I'm only seeing support for --package-db
2025-11-25 19:58:42 +0100 <hseg> unfortunately, it seems that setup.exe doesn't know to pick up on that file?
2025-11-25 19:58:11 +0100 <hseg> nope, the opposite -- it seems to be how it's communicating the package db info to the program it runs
2025-11-25 19:58:00 +0100 <geekosaur> (see `--package-env` in cabal)
2025-11-25 19:57:39 +0100 <geekosaur> usually you want that disabled, which I think is what cabal is doing
2025-11-25 19:57:10 +0100 <geekosaur> it's an environment file
2025-11-25 19:55:25 +0100 <hseg> right... what's GHC_ENVIRONMENT that I see cabal exec setting?
2025-11-25 19:55:15 +0100merijn(~merijn@host-vr.cgnat-g.v4.dfn.nl) (Ping timeout: 240 seconds)
2025-11-25 19:54:26 +0100 <geekosaur> but it would have potential issues since it would expose the store, which external commands could then modify without cabal knowing about it
2025-11-25 19:54:00 +0100Frostillicus(~Frostilli@pool-71-174-119-69.bstnma.fios.verizon.net)
2025-11-25 19:53:48 +0100 <geekosaur> mm, actually that's if you set it outside cabal. I don't know if it's possible to have cabal set it for commands it runs
2025-11-25 19:53:24 +0100Frostillicus(~Frostilli@pool-71-174-119-69.bstnma.fios.verizon.net) (Read error: Connection reset by peer)
2025-11-25 19:53:17 +0100 <hseg> it does
2025-11-25 19:53:10 +0100 <geekosaur> iirc cabal throws an error if you do that?
2025-11-25 19:50:54 +0100Googulator(~Googulato@2a01-036d-0106-4ad8-11e2-562e-881b-eb37.pool6.digikabel.hu) (Quit: Client closed)
2025-11-25 19:50:48 +0100Googulator56(~Googulato@2a01-036d-0106-4ad8-11e2-562e-881b-eb37.pool6.digikabel.hu)
2025-11-25 19:49:22 +0100 <hseg> is there some way of getting cabal to do that?
2025-11-25 19:49:14 +0100 <hseg> hrm. https://github.com/ndmitchell/shake/issues/267 and comments there suggest the test suite expects GHC_PACKAGE_PATH to be set properly in its environment