-
Notifications
You must be signed in to change notification settings - Fork 47
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
completely regular spaces and locally compact implies uniform #1331
Conversation
bd1b621
to
a074d00
Compare
ee04204
to
1057740
Compare
the change to also, |
Should "one point" by written "one-point" in the documentation? I was a bit suprised you chose to use "opc" as part of lemma |
I think And I definitely should not have exposed |
d55bb11
to
108e886
Compare
108e886
to
6e1f911
Compare
What about disabling the implicit parameter of |
I think we did make that change, and just never went back and removed al the extra
Removing the |
But instead of adding Arguments normal_space : clear implicits.
Arguments regular_space : clear implicits. you could also define Definition normal_space (T : topologicalType) :=
forall A : set T, closed A ->
filter_from (set_nbhs A) closure `=>` set_nbhs A.
Definition regular_space (T : topologicalType) :=
forall a : T, filter_from (nbhs a) closure --> a. to get the desired effect (then a user grepping the definitions will always found them fully applied as advertised in the header). |
Co-authored-by: affeldt-aist <[email protected]>
Co-authored-by: affeldt-aist <[email protected]>
We already have proofs about curry/uncurry which require locally compact and uniform. One one hand, these results are critical for homotopy theory. One the other hand they are badly phrased because locally compact implies uniform. That is what we prove here.
We follow the traditional proof, which is in 6 stages (2 of which were already done)
normal_separatorP
+uniform_separatorP
)sup {(weak_topology f) | f : X -> R is continuous}
)X
is in the weak topology from its OPC)However, it's non-forgetful to just globally assign a
Uniform
toX
. (And is legitimately dangerous. The nats with the discrete topology is hausdorff + locally compact, and does not have a unique uniformity).Instead we do something slightly more general, and formalize the "continuous functions into the reals" construction by with a type
completely_regular_uniformity
.So if you have a
crsX : completely_regular_space X
, then the typecompletely_regular_uniformity crsX
is uniform and satisfies(completely_regular_uniformity crsX: topologicalType) = X
. And coq is smart enough to exploit this equality, as seen incompletely_regular_regular
.Then it is enough to prove
locally_compact_completely_regular
, which do via a local non-forgetful inheritance, so it's ok.Checklist
CHANGELOG_UNRELEASED.md
Reference: How to document
Reminder to reviewers