К вопросу о заселении ⊥ при включенном type-in-type. Делается прямой реализацией парадокса Рассела
{-# OPTIONS --type-in-type --without-K #-}
module russell where
open import Data.Product
open import Relation.Binary.PropositionalEquality
open import Relation.Nullary
open import Data.Empty
-- a model of set theory
data U : Set where
set : (I : Set) → (I → U) → U
-- a set is regular if it doesn't contain itself
regular : U → Set
regular (set I f) = (i : I) → ¬ (f i ≡ set I f)
-- Russell's set: the set of all regular sets
R : U
R = set (Σ U regular) proj₁
-- R is not regular
R-nonreg : ¬ (regular R)
R-nonreg reg = reg (R , reg) refl
-- R is regular
R-reg : regular R
R-reg (x , reg) p = subst regular p reg (x , reg) p
-- contradiction
absurd : ⊥
absurd = R-nonreg R-reg
Взято отсюда:
http://www.reddit.com/r/compsci/comments/1a7g6q/a_question_on_set_classes_and_dependent_types/http://article.gmane.org/gmane.comp.lang.agda/5002http://hpaste.org/84029