Vyxal

lyxal

12:11 AM

sounds good to me

emanresu A

1:11 AM

Okay although I feel like I’ll mess it up

user

No there’s a button

You know the Update branch button on PRs? If you click on the drop down there’s an option to rebase

Vyxal Bot

2:15 AM

Steffan153 approved on PR #1151 (Vyxal/Vyxal)

Steffan153 merged PR #1151 (Vyxal/Vyxal) (Vyxal:fix-cartesian → Vyxal:main): Fix tests for `Ẋ` and `ÞẊ`

Steffan153 deleted branch Vyxal/fix-cartesian

Steffan153 approved on PR #1149 (Vyxal/Vyxal)

Steffan153 merged PR #1149 (Vyxal/Vyxal) (Vyxal:fix-depth → Vyxal:main): Make depth (`Þj`) work with empty children

Steffan153 deleted branch Vyxal/fix-depth

Steffan

@user We can just remove matrix reduce I think because it's just vectorised reduce

Vyxal Bot

Steffan153 merged PR #1148 (Vyxal/Vyxal) (Vyxal:fix-divide → Vyxal:main): Fix behavior of `/` and `Þ÷` when splitting strings/lists

Steffan153 deleted branch Vyxal/fix-divide

Steffan

!!/prod

Vyxal Bot

Vyxal-Bot opened PR #1152 (Vyxal/Vyxal) (Vyxal:main → Vyxal:production): Update Production (2022-06-24)

Steffan

132 commits. whew

lyxal

2:23 AM

That's a lotta commits

@user in fact you can make it the default somewhere iirc

Steffan

2:42 AM

pretty sure you can't

I looked it up and found people dying for the feature so it must not exist

Vyxal Bot

3:08 AM

Steffan153 created branch Vyxal/fix-from-base

Steffan153 opened PR #1153 (Vyxal/Vyxal) (Vyxal:fix-from-base → Vyxal:main): Fix from custom base

Steffan153 created branch Vyxal/fix-head-remove

Steffan153 opened PR #1154 (Vyxal/Vyxal) (Vyxal:fix-head-remove → Vyxal:main): Fix head remove with negative numbers

[Discord Relay]
cjquines#0001: you can set repo settings so that pull requests can't do merge

Steffan153 created branch Vyxal/fix-Ẇ

Steffan153 opened PR #1155 (Vyxal/Vyxal) (Vyxal:fix-Ẇ → Vyxal:main): Fix split and keep delimiter

Steffan153 created branch Vyxal/fix-surround

Steffan153 opened PR #1156 (Vyxal/Vyxal) (Vyxal:fix-surround → Vyxal:main): Fix surround

Vyxal Bot

3:45 AM

Steffan153 created branch Vyxal/fix-symmetric-difference

Steffan153 opened PR #1157 (Vyxal/Vyxal) (Vyxal:fix-symmetric-difference → Vyxal:main): Fix symmetric difference with string and number

Steffan153 created branch Vyxal/fix-unwrap

Steffan153 opened PR #1158 (Vyxal/Vyxal) (Vyxal:fix-unwrap → Vyxal:main): Fix unwrap

Steffan153 created branch Vyxal/fix-matrix-determinant

Steffan153 opened PR #1159 (Vyxal/Vyxal) (Vyxal:fix-matrix-determinant → Vyxal:main): Fix matrix determinant

Steffan153 created branch Vyxal/fix-untruth

Steffan153 opened PR #1160 (Vyxal/Vyxal) (Vyxal:fix-untruth → Vyxal:main): Fix untruth for empty lists

Steffan153 created branch Vyxal/fix-random-bits

Steffan153 opened PR #1161 (Vyxal/Vyxal) (Vyxal:fix-random-bits → Vyxal:main): Fix random bits

Steffan

4:03 AM

@VyxalBot yes

but you can't make it so that merge branch rebases

mold to shape is pretty broken (as seen in the great bug fix)

Vyxal Bot

[Discord Relay]
cjquines#0001: then… don't require merging main into the PR?

Steffan

how would you do that

oh man, we should just copy jelly

github.com/DennisMitchell/jellylanguage/blob/master/jelly/…

so clean and concise and small

Vyxal Bot

[Discord Relay]
cjquines#0001: in cases when github can rebase automatically, it just does it

[Discord Relay]
cjquines#0001: i'm constantly surprised we don't copy jelly more

Steffan

you can set it to do it automatically?

it will need changed some though, because LazyLists aren't mutable

it looks like you need an action to auto-update: github.com/marketplace/actions/auto-update

(which is fine with me)

Vyxal Bot

[Discord Relay]
cjquines#0001: uh

[Discord Relay]
cjquines#0001: what i mean is, pull requests don't need to be updated in order to be merged

[Discord Relay]
cjquines#0001: if the base branch doesn't require that branches be up to date before being merged
cjquines#0001: and if you turn off the setting to suggest updating branches

[Discord Relay]
cjquines#0001: <https://docs.github.com/en/pull-requests/collaborating-with-pull-requests/proposing-changes-to-your-work-with-pull-requests/keeping-your-pull-request-in-sync-with-the-base-branch>

Steffan

4:13 AM

suggesting updating branches is nice though

we might still edit the code and be affected

idk

Vyxal Bot

[Discord Relay]
cjquines#0001: actually, it's not about suggesting updating branches, i think?

[Discord Relay]
cjquines#0001: it's the setting "Require branches to be up to date before merging"

Steffan

so it will automatically rebase correctly when you merge?

Vyxal Bot

[Discord Relay]
cjquines#0001: yeah
cjquines#0001: it won't allow merging if there are conflicts

Steffan

it does mean though that if changes in main broke something in your PR, then the tests won't be ran with merged

so you might merge broken code

Vyxal Bot

[Discord Relay]
cjquines#0001: that is indeed the risk

[Discord Relay]
cjquines#0001: it's a question of how likely do you think it is that two PRs that don't conflict will break each other

Steffan

4:18 AM

yeah

anyway i gtg

!!/'night

Vyxal Bot

@Steffan o/

Steffan

10 PRs for people to review while I'm asleep

Vyxal Bot

[Discord Relay]
cjquines#0001: i've never actually seen this happen. i think we run checks on the main branch often enough that it's not a real risk

lyxal

5:39 AM

We used to copy Jelly more and that lead to the mess that was 2.5.3

Everything was in a single file and people just dev'd on main

Bad idea.

Vyxal Bot

[Discord Relay]
cjquines#0001: well, it's not like we're good at splitting things into files rn
cjquines#0001: `elements.py` *cough*

mathcat

9:59 AM

What are the digraphs for v3 again?

lyxal

10:27 AM

@mathcat ∆øÞ#

Those 4 characters

mathcat

lemme guess

∆ is constants, Þ is lists, ø is strings and # is other digraphs?

lyxal

No

∆øÞ are the same as now

# replaces the other digraphs

So actually you were quite close lol

mathcat

oh okay

lyxal

And k will continue to be for constants

mathcat

nice

lyxal

10:34 AM

# is really the only change in digraphs markers

user

11:03 AM

@Steffan ig sure

lyxal

12:17 PM

→ 9 PR approvals moved to Trash

lyxal

12:56 PM

@VyxalBot also, the problem with copying Jelly is that some things need to be adapted for the Vyxal ecosystem

for example, list molding is ripped from Jelly

but it's had to be modified to account for lazylists, immutability, etc

user

And it’s now borked

@VyxalBot i mean it doesn’t hurt to ensure everything is up to date

It doesn’t matter that much whether you merge or rebase i was just suggesting we try to rebase whenever possible

lyxal

@user is it though?

the mold test case says [[1,2,3], [[4], [], [6]]] -> [[1], [], [2]]

thain't whatever the hell this is

oh nvm it is

left an extra bracket lol

Vyxal Bot

1:29 PM

Lyxal created branch Vyxal/improve_mold

Lyxal opened PR #1162 (Vyxal/Vyxal) (Vyxal:improve_mold → Vyxal:main): Fix the molding functions

Lyxal created branch Vyxal/remove-a-testcase-for-§

Lyxal opened PR #1163 (Vyxal/Vyxal) (Vyxal:remove-a-testcase-for-§ → Vyxal:main): Remove a test case that was actually invalid

Lyxal opened issue #1164 in Vyxal/Vyxal: Ambiguities that arose from the test cases

lyxal

those aren't exactly bugs, but rather things that need a bit of discussion first

e.g. U: ["1",1] : ["1",1]

that could also be [1] because strings and integers can be considered equal under = but not under ⁼

user

1:45 PM

I feel like U should use ⁼

Vyxal Bot

cgccuser approved on PR #1163 (Vyxal/Vyxal)

lyxal

hey wait a second

U isn't broken at all

the test is

Vyxal Bot

cgccuser approved on PR #1162 (Vyxal/Vyxal): "I'm too smol brain to understand this code but the tests for this command seem to pass so LGTM"

user

lmao

I love when that happens

It is kinda hard to remember that the arguments need to be wrapped

lyxal

that's why the tests in the YAML for v3 are much nicer

user

1:51 PM

So...about that

Vyxal Bot

Lyxal created branch Vyxal/improve-uniquify

user

In the Scala version, I was thinking about having the tests in Scala itself

lyxal

not auto-generated by YAML?

user

Thus providing us with highlighting and the ability to write more complex tests right there

Yeah

Vyxal Bot

Lyxal opened PR #1165 (Vyxal/Vyxal) (Vyxal:improve-uniquify → Vyxal:main): Fix the test for uniquify

user

1:52 PM

I mean we could write a one-time script to port the YAML over to Scala

Vyxal Bot

cgccuser approved on PR #1161 (Vyxal/Vyxal)

lyxal

the only thing with that is that the YAML allows for people to just plug in tests without having to see if Scala compiles or anything

plus the current testing file is like 5k lines long

while the element list is shorter

user

Oh

lyxal

5k is a lie btw

it's 33k lines

user

Alright we can stick with the YAML, it'll be easier to not switch over ig

wtf

lyxal

1:54 PM

it's the YAML that's 6k

even with extra newlines removed, it's still 23.4k lines

(the test file)

user

Dear god

How did we get here?

lyxal

@Seggan wait what?

it should only return one item

user

Do we have a function to get the indices of a list? Like range(len(x)) but working for LazyLists

lyxal

would standard enumerate work?

user

Oh yeah

Also untruth is confusing me, the code makes a range to monadic_maximum(lhs) but the docs say that it's a range to len(lhs)

lyxal

2:00 PM

well it should be to len(lhs)

user

The docstring in the code says (any) -> [int(x in a) for x in range(max(a))]. Very sus

lyxal

no no it's to max

user

What's more useful?

lyxal

> Return a list with 1s at the (0-indexed) indices in a, and 0s elsewhere

at the indices in a

so not to len but to the maximal index

it's basically [a[x] = 1 for x in a; a[x] = 0 for x not in a]

weren't we going for rebase merges?

@lyxal I did rebase

2:12 PM

Oh not that one

ah

I meant rebasing instead of merging main into other branches when updating

lyxal

I did mean that one

user

lyxal

2:14 PM

oh

oh you mean like that

I thought you meant on the actual merge

user

22 hours ago, by user

So for PRs, I was thinking we could standardize to rebasing instead of merging main into feature branches, and squashing instead of merging directly

Wait so are you okay with squashing on the actual merge?

lyxal

Dec 29, 2021 at 3:46, by pxeger

Definitely don't squash when merging into production

user

I suggested squashing there so that main would only get 1 commit instead of every single commit from that branch (yesterday I pulled for the first time in a long time and there were like 300 commits)

lyxal

I think it's good to have the whole commit history avaliable

user

@lyxal Did he explain why?

Oh ok

lyxal

2:17 PM

Dec 29, 2021 at 3:46, by lyxal

@pxeger any reason in particular? (I merged like you suggested but I'm just curious)

never got an answer to that lol

user

:|

lyxal

found an explanation

github.com/Vyxal/Vyxal/pull/533#issuecomment-1002326867

which is an answer to the time you asked lol

user

lol

lyxal

Also, no, the surround PR won't introduce any ACE

because it's guaranteed the inputs are numbers

user

Cool cool

lyxal

2:29 PM

@VyxalBot that'll have to be for tomorrow when I know what I'm doing

Vyxal Bot

Steffan153 commented on PR #1162 (Vyxal/Vyxal): "you could copy this from jelly. it's just 9 lines github.com/DennisMitchell/jellylanguage/blob/master/jelly/…;

lyxal

@VyxalBot if only

but alas item = content.pop(0) doesn't exist on LinkedLists, and the way Dennis wrote it is kinda wack in that it modifies parameters in-place

Seggan

2:46 PM

@lyxal nvm i didnt read the specs

Seggan

3:56 PM

!!/issue Tooltips are broken again Uncaught SyntaxError: unterminated character class glyphSearch vyxal.pythonanywhere.com/:92 oninput vyxal.pythonanywhere.com/:1 online interpreter bug

Steffan

7:18 PM

@lyxal LinkedList lol bruh... anyway we could just convert it to a list

we're traversing through the whole thing non-lazily anyway

user

7:32 PM

→ 38 messages moved to Trash

Steffan

8:17 PM

why is vyxal so so so SO slow

just a simple task like cartesian power or product can take more than ten seconds, while Jelly and 05AB1E output instantly

I don't get it

and Jelly is Python too

emanresu A

Hmm

mathcat

~~use myxal~~

Does jelly use sympy too?

emanresu A

Connection speeds and lazylists I guess

Steffan

or take my latest Vyxal answer. give it [2,1,2,4,4] and it times out. now give that to Jelly, which uses exactly the same algorithm, and watch it output instantly

mathcat

or is it because of pythonanywhere

emanresu A

8:21 PM

Try it locally?

mathcat

^

Steffan

is pythonanywhere that much better than TIO?

just a sec

mathcat

tio has its own domain, so it might be because of that