| 6 Jun 2026 |
 hellwolf |
- I checked out the latest https://github.com/well-typed/optics
- ran cabal test optics
- with ghc910, passes
- with ghc912, 1 failed.
| 14:01:08 |
 alexfmpe | Huh same seed? Maybe it's some ghc bug or some sketchy optimizations thing | 14:03:11 |
| alexfmpe | Definitely file issue | 14:03:24 |
| hellwolf | same seed.
reported.
Fwiw, the haskell-updates build produces 2 more errors. | 14:09:45 |
| 7 Jun 2026 |
 sterni | Let's hope this goes away with GHC 9.12.5-rc1 :) | 19:47:21 |
|  WorldsGoneMad joined the room. | 20:34:18 |
 MangoIV | I’ll try tomorrow. | 21:18:06 |
| MangoIV | Do you have a link to the actual errors (this is missing the errors I think)? Otherwise I can just reproduce tomorrow myself | 21:19:11 |
| 8 Jun 2026 |
| hellwolf | Redacted or Malformed Event | 09:39:22 |
| hellwolf | https://github.com/well-typed/optics/issues/545 | 09:41:08 |
| hellwolf | MangoIV: I am still in zurihac, fwiw. In the work room near aula auditorium. | 09:41:33 |
| hellwolf | is it a ghc bug? | 10:17:26 |
| hellwolf | I am now testing with the tip of ghc's "9.12" branch | 12:50:57 |
| hellwolf | oh, no
<no location info>: error:
panic! (the 'impossible' happened)
GHC version 9.12.4.20260607:
ASSERT failed!
in_scope InScope {ds_aP9t $dMonoid_s26Ke $dFunctor_s26Kg
$dApplicative_s26Ki lvl_s26Ox lvl_s26OA lvl_s26Wj lvl_s26Wk
lvl_s26Wl lvl_s26Wm}
tenv [a1gar :-> f_a1ga3, a1gas :-> t_a1ga4, a1gat :-> a_a1ga5,
a1gau :-> r_a1ga6]
tenvFVs {f_a1ga3, t_a1ga4, a_a1ga5, r_a1ga6}
cenv []
cenvFVs {}
tys []
cos [<t_a1gas a_a1gat -> Traversed f_a1gar r_a1gau>_R
%<'Many>_N ->_R Sym (N:Forget
<Traversed f_a1gar r_a1gau>_R
<ZonkAny 1>_P
<t_a1gas a_a1gat>_R
<()>_P)]
Call stack:
CallStack (from HasCallStack):
assertPpr, called at compiler/GHC/Core/TyCo/Subst.hs:680:5 in ghc-9.12.4.20260607-inplace:GHC.Core.TyCo.Subst
checkValidSubst, called at compiler/GHC/Core/TyCo/Subst.hs:848:17 in ghc-9.12.4.20260607-inplace:GHC.Core.TyCo.Subst
substCo, called at compiler/GHC/Core/Subst.hs:261:39 in ghc-9.12.4.20260607-inplace:GHC.Core.Subst
substExpr, called at compiler/GHC/Core/Subst.hs:268:37 in ghc-9.12.4.20260607-inplace:GHC.Core.Subst
substExpr, called at compiler/GHC/Core/Subst.hs:272:37 in ghc-9.12.4.20260607-inplace:GHC.Core.Subst
substExpr, called at compiler/GHC/Core/Subst.hs:272:37 in ghc-9.12.4.20260607-inplace:GHC.Core.Subst
substExpr, called at compiler/GHC/Core/Subst.hs:306:28 in ghc-9.12.4.20260607-inplace:GHC.Core.Subst
substBind, called at compiler/GHC/Core/Subst.hs:274:44 in ghc-9.12.4.20260607-inplace:GHC.Core.Subst
substExpr, called at compiler/GHC/Core/Subst.hs:272:37 in ghc-9.12.4.20260607-inplace:GHC.Core.Subst
substExpr, called at src/Test/Inspection/Core.hs:609:18 in inspection-testing-0.6.3-152826bba0790f422bcf299ee350f11f2d91c1b0e54a3e3d07dd75e90d5a68a4:Test.Inspection.Core
Please report this as a GHC bug: https://www.haskell.org/ghc/reportabug
| 12:54:12 |
| 10 Jun 2026 |
|  magnolia_mayhem changed their profile picture. | 12:24:25 |
| 11 Jun 2026 |
|  @tillerino:matrix.org left the room. | 07:33:26 |
|  lav joined the room. | 23:50:32 |
| 13 Jun 2026 |
|  klea (she/her) joined the room. | 21:38:54 |
| 18 Jun 2026 |
|  whispers [& it/fae] changed their display name from whispers [& it/fae] to meow meow. | 18:46:03 |
| whispers [& it/fae] changed their display name from meow meow to whispers [& it/fae]. | 19:11:51 |
| 19 Jun 2026 |
|  Prince joined the room. | 22:31:04 |
| 22 Jun 2026 |
|  @albertlarsan68:albertlarsan.fr left the room. | 17:39:10 |
| 23 Jun 2026 |
|  instinctive joined the room. | 18:39:39 |
| 24 Jun 2026 |
 Janus | i saw https://github.com/NixOS/nixpkgs/pull/521260 and i am wondering how the timing is supposed to be. does it mean that staging will update before the LTS, and then when LTS 25 gets released, master will eventually pick up the final LTS 25 changes in later batches? | 21:29:17 |
| alexfmpe | there's some knot tying involved | 22:04:33 |
| alexfmpe | nightly promotion to LTS is blocked on a GHC 9.12.x without critical bugs | 22:04:48 |
| alexfmpe | 9.12.5 RCs are using stackage/nightly and nixpkgs/haskell-updates as a test bed | 22:05:32 |
| alexfmpe | once 9.12.5 proper gets released, stackage can turn nightly into LTS 25 and nixpkgs can follow suit | 22:06:55 |
| alexfmpe | once 10.0 and reinstallable base comes around we probably don't need the knot tying anymore | 22:07:39 |
| alexfmpe | in haskell-updates, both http-reverse-proxy and jsaddle-warp test suites fail with
GHC.Internal.Event.Thread.getSystemTimerManager: the TimerManager requires linking against the threaded runtime
Encountered an error trying to make a request: socket hang up
| 22:41:31 |
