Zulip Chat Archive
Stream: Is there code for X?
Topic: Typeclass for indiscrete topology
Aaron Liu (Apr 14 2025 at 21:10):
Something like this:
import Mathlib
class IndiscreteTopology (X : Type*) [t : TopologicalSpace X] : Prop where
eq_top : t = ⊤
Do we have it? If we don't, do we want to have it?
Kevin Buzzard (Apr 14 2025 at 21:17):
Is it useful? I think the discrete topology is a lot more common.
Aaron Liu (Apr 14 2025 at 21:19):
Recently Just now I found myself wanting this, because a lot of theorems are true assuming the indiscrete topology. It's equivalent to Subsingleton (SeparationQuotient X) and I'm in the middle of generalizing some theorems to non-T0 spaces.
Aaron Liu (Apr 14 2025 at 21:23):
Maybe I'll come back to this if I need it again but for now I found a way around.
Edward van de Meent (Apr 14 2025 at 21:34):
a quick loogle search says there is no such thing in mathlib... the only relevant things i could find were docs#induced_const and docs#induced_top
Edward van de Meent (Apr 14 2025 at 21:35):
my query was loogle TopologicalSpace, ?t = ⊤
Yaël Dillies (Apr 14 2025 at 21:37):
@loogle (⊤ : TopologicalSpace _)
loogle (Apr 14 2025 at 21:37):
:search: continuous_top, nhds_top, and 11 more
Eric Wieser (Aug 06 2025 at 22:30):
I found myself wanting this too; we have a few lemmas using ∃ x, ‖x‖ ≠ 0 for seminormed spaces which end up with API duplication for nontrivial normed spaces.
Eric Wieser (Aug 06 2025 at 22:34):
Oh, I guess I wanted NotIndiscrete
Notification Bot (Aug 06 2025 at 23:32):
2 messages were moved from this topic to #Is there code for X? > Typeclass for nontrivial topology by Eric Wieser.
Notification Bot (Aug 06 2025 at 23:33):
A message was moved here from #Is there code for X? > Typeclass for nontrivial topology by Eric Wieser.
Eric Wieser (Aug 07 2025 at 21:24):
Though in fact I'd quite like both, as a nicer way to spell cases subsingleton_or_nontrivial (SeparationQuotient X)
Last updated: Dec 20 2025 at 21:32 UTC