2025/06/03

Newest at the top

2025-06-03 10:27:00 +0200sabathan2(~sabathan@amarseille-159-1-12-107.w86-203.abo.wanadoo.fr)
2025-06-03 10:26:50 +0200notzmv(~daniel@user/notzmv) notzmv
2025-06-03 10:26:35 +0200sabathan2(~sabathan@amarseille-159-1-12-107.w86-203.abo.wanadoo.fr) (Remote host closed the connection)
2025-06-03 10:26:07 +0200tromp(~textual@2001:1c00:3487:1b00:7d34:d696:3458:218b) (Quit: My iMac has gone to sleep. ZZZzzz…)
2025-06-03 10:25:49 +0200notzmv(~daniel@user/notzmv) (Ping timeout: 245 seconds)
2025-06-03 10:25:39 +0200 <Leary> `-fdiagnostics-as-json` should be stable enough so long as you only check the fields you actually need to care about.
2025-06-03 10:21:05 +0200sabathan2(~sabathan@amarseille-159-1-12-107.w86-203.abo.wanadoo.fr)
2025-06-03 10:20:38 +0200sabathan2(~sabathan@amarseille-159-1-12-107.w86-203.abo.wanadoo.fr) (Remote host closed the connection)
2025-06-03 10:20:26 +0200sawilagar(~sawilagar@user/sawilagar) sawilagar
2025-06-03 10:19:58 +0200 <dminuoso> Or use ghc-lib with tight bounds.
2025-06-03 10:19:44 +0200 <dminuoso> One thing I would keep in mind, if you do not use ghc-lib and invoke ghc as a binary instead, you should tightly control the version as error output can change across GHC versions.
2025-06-03 10:19:39 +0200fp(~Thunderbi@hof1.kyla.fi) (Ping timeout: 272 seconds)
2025-06-03 10:18:39 +0200 <dminuoso> Loads of options
2025-06-03 10:18:30 +0200 <dminuoso> hellwolf: Just to give more perspectives, GHC does these golden tests by having a python based test driver https://gitlab.haskell.org/ghc/ghc/-/tree/master/testsuite/driver and then keeping the testcases + expected stderr next to each other: https://gitlab.haskell.org/ghc/ghc/-/tree/master/testsuite/tests/typecheck/should_fail
2025-06-03 10:16:13 +0200sabathan2(~sabathan@amarseille-159-1-12-107.w86-203.abo.wanadoo.fr)
2025-06-03 10:15:52 +0200sabathan2(~sabathan@amarseille-159-1-12-107.w86-203.abo.wanadoo.fr) (Remote host closed the connection)
2025-06-03 10:15:27 +0200Lycurgus(~juan@user/Lycurgus) (Quit: irc.renjuan.org (juan@acm.org))
2025-06-03 10:12:46 +0200 <lortabac> https://github.com/simonmichael/shelltestrunner
2025-06-03 10:10:56 +0200 <lortabac> https://github.com/lortabac/parameters/blob/master/.github/workflows/haskell.yml#L53
2025-06-03 10:10:17 +0200 <hellwolf> I wonder if you use Unsatisfiable, would you be able to test if instance selection falls to the unsatisiable one.
2025-06-03 10:09:44 +0200 <lortabac> I have used it in one of my projects for this exact purpose
2025-06-03 10:09:15 +0200 <lortabac> there is also shelltest
2025-06-03 10:08:59 +0200sabathan2(~sabathan@amarseille-159-1-12-107.w86-203.abo.wanadoo.fr)
2025-06-03 10:08:35 +0200sabathan2(~sabathan@amarseille-159-1-12-107.w86-203.abo.wanadoo.fr) (Remote host closed the connection)
2025-06-03 10:07:54 +0200 <lortabac> **some** tests can also be performed with the Data.Type.Equality machinery
2025-06-03 10:07:05 +0200machinedgod(~machinedg@d108-173-18-100.abhsia.telus.net) machinedgod
2025-06-03 10:05:44 +0200 <dminuoso> It's a bit less tightly coupled and might be easier to use.
2025-06-03 10:04:50 +0200 <dminuoso> hellwolf: Alternatively you could invoke GHC as a binary and use -fdiagnostics-as-json
2025-06-03 10:04:11 +0200sabathan2(~sabathan@amarseille-159-1-12-107.w86-203.abo.wanadoo.fr)
2025-06-03 10:03:49 +0200sabathan2(~sabathan@amarseille-159-1-12-107.w86-203.abo.wanadoo.fr) (Remote host closed the connection)
2025-06-03 10:03:33 +0200 <dminuoso> What do you mean you would be surprised?
2025-06-03 10:03:11 +0200 <hellwolf> Okay, that changes things. I will look into it some day. I would still be surprised it's not how one should generally test type errors.
2025-06-03 09:58:18 +0200sabathan2(~sabathan@amarseille-159-1-12-107.w86-203.abo.wanadoo.fr)
2025-06-03 09:57:56 +0200sabathan2(~sabathan@amarseille-159-1-12-107.w86-203.abo.wanadoo.fr) (Read error: Connection reset by peer)
2025-06-03 09:57:01 +0200trickard(~trickard@cpe-51-98-47-163.wireline.com.au)
2025-06-03 09:55:48 +0200merijn(~merijn@77.242.116.146) merijn
2025-06-03 09:52:11 +0200sabathan2(~sabathan@amarseille-159-1-12-107.w86-203.abo.wanadoo.fr)
2025-06-03 09:51:47 +0200sabathan2(~sabathan@amarseille-159-1-12-107.w86-203.abo.wanadoo.fr) (Remote host closed the connection)
2025-06-03 09:51:10 +0200trickard(~trickard@cpe-51-98-47-163.wireline.com.au) (Read error: Connection reset by peer)
2025-06-03 09:50:04 +0200sabathan2(~sabathan@amarseille-159-1-12-107.w86-203.abo.wanadoo.fr)
2025-06-03 09:49:35 +0200sabathan2(~sabathan@amarseille-159-1-12-107.w86-203.abo.wanadoo.fr) (Remote host closed the connection)
2025-06-03 09:46:32 +0200trickard(~trickard@cpe-51-98-47-163.wireline.com.au)
2025-06-03 09:45:56 +0200xff0x(~xff0x@fsb6a9491c.tkyc517.ap.nuro.jp)
2025-06-03 09:44:31 +0200 <dminuoso> (ghc-lib has the equivalent exports, but for some reason the haddock is missing on hackage)
2025-06-03 09:44:18 +0200trickard(~trickard@cpe-51-98-47-163.wireline.com.au) (Ping timeout: 248 seconds)
2025-06-03 09:44:15 +0200 <dminuoso> https://hackage.haskell.org/package/ghc-9.12.1/docs/GHC.html#v:parseModule and followingn typecheckModule
2025-06-03 09:44:07 +0200sabathan2(~sabathan@amarseille-159-1-12-107.w86-203.abo.wanadoo.fr)
2025-06-03 09:43:38 +0200sabathan2(~sabathan@amarseille-159-1-12-107.w86-203.abo.wanadoo.fr) (Remote host closed the connection)
2025-06-03 09:42:27 +0200sawilagar(~sawilagar@user/sawilagar) (Read error: Connection reset by peer)
2025-06-03 09:42:21 +0200tromp(~textual@2001:1c00:3487:1b00:7d34:d696:3458:218b)