-
Notifications
You must be signed in to change notification settings - Fork 49
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
splitting function space topologies #1166
splitting function space topologies #1166
Conversation
I made a first pass. A few things (doc) were not moved, a few removed things were not documented in the changelog, and I formatted a bit the documentation and suppressed a couple of parentheses here and there. I left a few comments mostly documentation that we'd better address now. Yet, a very welcome PR! |
As a last check I mostly checked the doc for typos. |
f982af8
to
81a3ce4
Compare
I added two NBs and force-pushed to merge asap. |
Motivation for this change
Three important changes here.
topology.v
file is huge, and kinda unpleasant for development. This splits all the function space topology stuff into a separate file. Note thatprod_topology
(notU
*V
, butforall i : U, K i
is a function space. So all that stuff, including tychonoff's theorem, lives infunction_spaces.v
.->
. It's a bit arbitrary to choose the topology of uniform convergence, and it was causing other bugs. Assigning topologies to aliases of->
likecompact-open
had to be done in modules to avoid some HB errors. This cleans all that up. It also introducesArrowAsX
modules that can locally assign a canonical topology to->
.pointwise_fun
, which was a layer of indirection that I believe is no longer serving a purpose. It was an alias forprod_topology (fun (_:U) => V)
. But should have all the same canonicals. Seems better to just use the dedicated notation{ptws, U -> V}
.A few other notes.
Filter F
into the context. Not a big deal, but a little surprising. All the instances areGlobal
, so I can't tell why the inference would be different. I don't consider this blocking, as it's a very minor inconvenience.topology.v
. Probably worth a quick scan to double check.Checklist
CHANGELOG_UNRELEASED.md
Reference: How to document
Reminder to reviewers