Zulip Chat Archive

Stream: new members

Topic: 0 < ε ≤ 0.5 (joke)


Eric Wieser (Apr 27 2025 at 11:52):

Edward van de Meent said:

has (hε : 0 < ε ≤ 0.5) ever been valid notation? i don't think so, right?

import Mathlib

instance : OfScientific Prop where
  ofScientific m s e :=
    let q :  := OfScientific.ofScientific m s e
    if q  0 then
      False
    else if 1  q then
      True
    else
      Nat = Int


-- `≤ 0.5` means "at most as true as undecidable"
example (ε : Real) ( : ε = 0) :
    (0 < ε)  0.5 := by
  aesop

is close

Kevin Buzzard (Apr 27 2025 at 13:26):

In 2018, a<b<ca<b<c would hang Lean 3 if they were real numbers, as my students would sometimes discover; it would get parsed as (a<b)<c and then Lean would hang looking for coercions from Prop to real or real to Prop. I think this was before the real number rewrite.


Last updated: May 02 2025 at 03:31 UTC