A RetroSearch Logo

Home - News ( United States | United Kingdom | Italy | Germany ) - Football scores

Search Query:

Showing content from https://github.com/agda/agda/issues/7803 below:

Internal error with Path and with-abstraction · Issue #7803 · agda/agda · GitHub

The following code results in an internal error:

{-# OPTIONS --cubical #-}
open import Agda.Builtin.Cubical.Path

p : SetSet
p i with _
... | _ = Set
An internal error has occurred. Please report this as a bug.
Location of the error: __IMPOSSIBLE__, called at src/full/Agda/TypeChecking/Telescope/Path.hs:71:14 in Agda-2.8.0-2UHLkJpHripCLhmKYbjPID:Agda.TypeChecking.Telescope.Path

which is located at:

_ -> __IMPOSSIBLE__ -- 2024-10-07 Andreas, issue #7413 -- Andrea writes in https://github.com/agda/agda/issues/7413#issuecomment-2396146135 -- -- I believe this is actually impossible at the moment -- unless generalized Path types were implemented while I wasn't looking: -- -- telePathPi only does this check if there's a boundary, -- which should only be introduced by a PathP copattern, -- which then should ensure the result type is in Type lvl -- for some lvl that does not depend on on the interval -- variable of the path. -- -- WAS: generic error with message -- text "The type is non-fibrant or its sort depends on an interval variable" <+> prettyTCM (unAbs b)

with a comment referencing

#7413

.

#7539

changed it from a

GenericError

to an

__IMPOSSIBLE__

.

So the code above was a reproducer for the error in the issue, since 2.7.0.1 used to throw:

The type is non-fibrant or its sort depends on an interval variable
_4 (i = i) → Set₁

RetroSearch is an open source project built by @garambo | Open a GitHub Issue

Search and Browse the WWW like it's 1997 | Search results from DuckDuckGo

HTML: 3.2 | Encoding: UTF-8 | Version: 0.7.4