Zulip Chat Archive
Stream: maths
Topic: Notation for `nhds_within`
Yury G. Kudryashov (Jul 30 2020 at 05:48):
What do you think about (a) introducing a notation for nhds_within
? E.g., ๐[s] a
? (b) using notation
instead of def
?
Removing def
makes all lemmas about inf_principal
work without extra dunfold
.
Yury G. Kudryashov (Jul 30 2020 at 05:50):
@Patrick Massot @Sebastien Gouezel :up:
Johan Commelin (Jul 30 2020 at 07:28):
You mean that you want to abolish the def nhds_within
and replace it with notation?
Johan Commelin (Jul 30 2020 at 07:28):
Does it help if you make nhds_within
reducible, or an abbreviation
?
Yury G. Kudryashov (Jul 30 2020 at 07:43):
AFAIR, either simp
or rw
won't be able to rewrite on lemmas about *_inf_principal
in this case.
Patrick Massot (Jul 30 2020 at 09:13):
I like notations.
Yury G. Kudryashov (Dec 09 2021 at 12:41):
I think that we should have notation for ๐[Ici x] x
, ๐[Iic x] x
, ๐[Ioi x] x
, ๐[Iio x] x
, and ๐[{x}แถ] x
. What about ๐แตฃ x
, ๐โ x
, ๐แตฃ' x
, ๐โ' x
, and ๐' x
? My main concern is that we need to type x
twice which is bad, e.g., in ๐[{(2 * k + 1) * ฯ / 2}แถ] ((2 * k + 1) * ฯ / 2)
.
Eric Wieser (Dec 09 2021 at 13:24):
Would ๐[โฅ] x
, ๐[โค] x
, ๐[>] x
, ๐[<] x
, and ๐[แถ] x
be valid notation?
Floris van Doorn (Dec 09 2021 at 13:29):
those would be valid notation yeah, but they are pretty confusing to me (note that the inequality is to the left of the x
...)
Yury G. Kudryashov (Dec 09 2021 at 19:02):
Anatole Dedecker (Dec 09 2021 at 20:57):
What about ๐โ
instead of ๐'
? It has a clearer meaning to me, but maybe that's too annoying to type (note that it's \*
not *
)
Yury G. Kudryashov (Dec 13 2021 at 06:21):
So, which notation should we use? I have #10694 with ๐แตฃ x
, ๐โ x
, ๐แตฃ' x
, ๐โ' x
, and ๐' x
but I can switch to some other notation easily.
Sebastien Gouezel (Dec 13 2021 at 07:20):
I am very happy with your notations. The only reason I haven't reviewed it yet is that the PR text still reads I'll add notation to module docs later today.
:-)
Yury G. Kudryashov (Dec 13 2021 at 07:57):
Done.
Eric Wieser (Dec 13 2021 at 08:44):
(no objection from me, I only suggested the other notation to see if anyone else preferred it)
Yury G. Kudryashov (Dec 13 2021 at 08:49):
@Floris van Doorn I don't find @Eric Wieser 's notation confusing: ๐[<] x = ๐[Iio x] x = ๐[{y | y < x}] x
agrees with x
being on the right of the inequality sign.
Yury G. Kudryashov (Dec 13 2021 at 08:51):
Let me create a poll (closing 24 h from now).
Yury G. Kudryashov (Dec 13 2021 at 08:53):
/poll Notation for ๐[Ici x] x
, ๐[Iic x] x
, ๐[Ioi x] x
, ๐[Iio x] x
, and ๐[{x}แถ] x
๐แตฃ x
, ๐โ x
, ๐แตฃ' x
, ๐โ' x
, and ๐' x
;
๐[โฅ] x
, ๐[โค] x
, ๐[>] x
, ๐[<] x
, and ๐[แถ] x
;
๐[โฅ] x
, ๐[โค] x
, ๐[>] x
, ๐[<] x
, and ๐[โ ] x
Floris van Doorn (Dec 13 2021 at 10:07):
Oh, my bad. I had it backwards indeed.
Eric Wieser (Dec 13 2021 at 10:12):
One more option would be to have ๐'[r] x
as notation for ๐'[{y | r y x}] x
and then ๐'[(โฅ)] x
etc are supported automatically.
Eric Wieser (Dec 13 2021 at 10:13):
But that's more verbose and unfolds to something less nice.
Yury G. Kudryashov (Dec 14 2021 at 08:56):
Switched to ๐[โฅ] x
, ๐[โค] x
, ๐[>] x
, ๐[<] x
, and ๐[โ ] x
Yury G. Kudryashov (Dec 14 2021 at 08:56):
#10694 is ready for review
Last updated: Dec 20 2023 at 11:08 UTC