Vyxal

emanresu A

01:48

@mathcat We should probably do magnitude or something

lyxal

04:31

Since you last saw the codepage, I've shifted some things around, and added some more greek characters

  0123456789ABCDEF
  ----------------
0|¡¿¤¢€©®§¶æð▲▼ƒ⟨⟩
1|λƛΩΛµ∆øÞŒʑʒʓ×÷£¥
2| !"#$%&'()*+,-./
3|0123456789:;<=>?
4|@ABCDEFGHIJKLMNO
5|PQRSTUVWXYZ[\]^_
6|`abcdefghijklmno
7|pqrstuvwxyz{|}~¦
8|ȦḂĊḊĖḞĠḢİĿṀṄȮṖṘṠ
9|ṪẆẊẎŻȧḃċḋėḟġḣŀṁṅ
A|ȯṗṙṡṫẇẋẏż≤≥≠⁼→←‡
B|₀₁₂₃₄₅₆₇₈₉«»⌐₩₳⊻
C|⁰¹²³⁴⁵⁶⁷⁸⁹⁺⁻⁾√∑Π
D|ẠḄḌẸḤỊḲḶṂṆỌṚṢṬạḅ
E|ḍẹịḷṃọṛṣṭ±□‛“¯≈ı
F|‹›βεδζξστχω↑↓∥∦≡

Example program:

₁ƛ35fḊkFḢḍṀ

emanresu A

Honestly I think a lot of stuff should stay the same if possible

lyxal

₁ƛ35fḊkFḢḍṀ
₁           ⍝ Push 100 to the stack
 ƛ          ⍝ To each item n in the range [1, 100]:
  35f       ⍝   Push the digits of 35
     Ḋ      ⍝   And push whether n is divisible by each
      kFḢ   ⍝   ["Fizz", "Buzz"]
         ḍ  ⍝   Dot product of the two (a shortcut for ×∑)
          Ṁ ⍝   Dyadic maximum

@emanresuA And it mostly is, except for a few things I always wanted to fix

emanresu A

Hm okay

(like ↑↓?)

lyxal

Those are modifiers now

Here's some of the key changes:

- Dotted uppercase comes before dotted lowercase

PyGamer0

@lyxal wait is that the APL comment

emanresu A

04:38

IMO # is a good comment character for polyglotting and source layout

lyxal

@PyGamer0 yes, because as suggested in an issue, comments won't be part of the codepage

PyGamer0

i mean lol flax
has the same character

lyxal

Yes, but ⍝ isn't in the codepage

it's merely a utility for comments

If you submit ⍝ in a program, you'd have to use UTF-8 scoring

@emanresuA sounds like a flag idea

emanresu A

... I guess

PyGamer0

@lyxal a flag to change comment chars o_O

lyxal

04:42

@emanresuA but you can still polyglot though

#<element><rest_of_code>

emanresu A

wdym?

lyxal

# is a modifier

emanresu A

What does it do?

lyxal

so it just modifies the next item

@emanresuA to be determined

emanresu A

Hmm okay

lyxal

04:44

It's in my list of 34 characters to be modifiers

¡¿¤¢€©®§æðƒ"#)/\]}~¦βεδζξστχω↑↓∥∦≡

emanresu A

Modifiers are postfix right?

lyxal

In version 3, most probably

I say most probably because I'm yet to fully decide whether or not auto-closing can be accounted for with postfix

e.g. λ<lambda_code>v wouldn't vectorise the whole lambda

emanresu A

True...

lyxal

so I had the idea of λv<lambda_code>

but I don't know how easy that is to parse

emanresu A

Or just <lambda-code>λv

lyxal

04:48

no

lambdas are structures

emanresu A

Hmm okay

lyxal

and besides then how would it know where the lambda starts?

the <lambda_code> is to be replaced entirely

the < and > are part of the placeholder

λv... should paint a clearer picture

emanresu A

I know, but didn't you want a modifier that parsed until start of line?

lyxal

yes, and that'll be implemented

but what about those edge cases where you can't use that modifier?

but you still need an auto-closed structure

this method would also work for other structures

[v<onTrue>|<onFalse>

(<variableName>|v...

emanresu A

True...

lyxal

04:57

I never finished my key changes list for the codepage, so here it is in its entirety:

- Uppercase dotted letters now come before lowercase dotted letters
- `₉` is now included in the subscript numbers range.
- There's a full range of superscript numbers too.
- Bitshifting stuff is now monospaced.
- A lot of the non-monospaced characters have been swapped out.
- More greek letters!
- A fifth lambda type can now be used.
- The global array commands are no longer fractions.
- Downdot letters exist. Now you can updot _and_ downdot.
- There's now a different srting type for dictionary compression.

emanresu A

@lyxal What's the fifth lambda type?

lyxal

@emanresuA reduce by at this stage

emanresu A

Okay

lyxal

Is it obvious that ∥ will be parallel apply?

PyGamer0

05:24

@lyxal 35 modifiers o_O

oh wait i have 20 quicks

@lyxal yes

emanresu A

@lyxal Now that you've told me :P

lyxal

@emanresuA you didn't recognise that it's the mathematical symbol for parallel?

PyGamer0

@lyxal well i thought || is the symbol

lyxal

@PyGamer0 it is.

∥ is a single unicode character

|| != ∥

PyGamer0

yes but pretty sure there was a vertial version of that?

lyxal

05:28

that is vertical

PyGamer0

no it isnt

lyxal

yes, it is

PyGamer0

its oblique (frick you fonts)

for me

lyxal

PyGamer0

∥ omg it is vertical

lyxal

05:29

is vertical

PyGamer0

whytf was it rendering oblique

lyxal

admittedly that's a screenshot from vs code

also, you just gotta zoom in

PyGamer0

and i assume you use consolas

lyxal

^ is from chat

with a little zoom

PyGamer0

^ it looks like that for me

lyxal

05:32

yikes

that's a big F from me

even with my dark mode extension and all the userscripts I have turned off, it still renders properly

Ugh light mode chat is an ugly sight to behold

mathcat

09:35

@lyxal amogus comment!!!

@lyxal which userscript are you using?

lyxal

09:47

@mathcat I'm using a chrome extension

I'll get it for you when I'm next at my computer

mathcat

ok, thanks

Idea: make another context variable, which acts as a counter for loops, maps, etc.

Is an element worth that?

mathcat

10:22

What is the fastest way to join like A,B,A (without repeating A)?

!!/run 2 1J 2J

Vyxal Bot

@mathcat 212

mathcat

!!/run `abc``def`J`abc`J

Vyxal Bot

@mathcat abcdefabc

lyxal

Use over?

Ȯ

@mathcat I was thinking making that m

mathcat

nice thanks

@lyxal yeah, currently it's 2 bytes for every usage

somehow my program is actually working

DialFrost

10:34

hai

mathcat

@mathcat That's about (once in a blue moon)¹⁰

DialFrost

im start learning in around 2 days time @lyxal

lyxal

Very good

@mathcat github.com/darkreader/darkreader

oop wrong reply

but right person

mathcat

looks scary

I'll get used to it

lyxal

so used to it that you'll not be able to use light mode SE lol

Vyxal Bot

10:44

Lyxal created branch Vyxal/2.8.3-version-bump

Lyxal opened PR #730 (Vyxal/Vyxal) (Vyxal:2.8.3-version-bump → Vyxal:main): Version 2.8.3

lyxal

just a simple version bump so that the pypi package update doesn't error

Vyxal Bot

Not-Lyxal approved on PR #730 (Vyxal/Vyxal)

mathcat

when was it released?

lyxal

in maybe half an hour

mathcat

oh, it's not released yet?

lyxal

10:46

I have to do the version bump before writing the release notes

@mathcat exactly

I usually just force merge it with alts

but you can approve it if you want

Vyxal Bot

Tams-Tams approved on PR #730 (Vyxal/Vyxal): "***I can approve***"

Lyxal merged PR #730 (Vyxal/Vyxal) (Vyxal:2.8.3-version-bump → Vyxal:main): Version 2.8.3

Lyxal deleted branch Vyxal/2.8.3-version-bump

lyxal

Good

now I can write the release notes

which is really just clicking a button

and fixing any mistakes

very helpful button

Vyxal Bot

Version 2.8.3

5

lyxal

there we go

with a little luck, the pypi will be updated properly and there won't be any workflow issues

good, v2.8.3 is live on PyPi

Vyxal Bot

Lyxal created branch Vyxal/v3-planning-documents

lyxal

10:53

Very important: Do not merge the v3-planning-documents branch into main - that'd delete everything, and we don't want to do that.

4

PyGamer0

@lyxal so everything we plan goes there?

and will v3 be interpreted?

lyxal

@PyGamer0 that's correct

and it will be interpreted

v3 elements.yaml

PyGamer0

wooo interpretation

@lyxal wait the should be rewritten

so we are basically rewriting the entire thing

lyxal

not entirely

github.com/Vyxal/Vyxal/blob/experimental

some stuff can be taken from there

the experimental branch was my first attempt at version 3 planning way back in June 2021 when everything wasn't broken

PyGamer0

@lyxal 404

lyxal

11:02

github.com/Vyxal/Vyxal/tree/experimental

PyGamer0

@lyxal question: will classes support inheritance?

lyxal

no. Only composition

unless someone can present a reason why inheritance is absolutely necessary, I don't think so.

PyGamer0

@lyxal wdym

lyxal

pxeger presented reasons why not to have inheritance for classes

PyGamer0

@lyxal oh boy i got a weird idea: implement only the structures and the most necessary commands and the rest can be made from those commands

kinda like BQN

@lyxal i saw that, i want to know the meaning of class composition

lyxal

11:05

classes containing instances of other classes to mimic inhertiance

PyGamer0

oh

lyxal

besides, it makes class implementation easier

PyGamer0

has everyone agreed to use scala? or is there a chance we will use python?

lyxal

scala is our language of choice

PyGamer0

ok

lyxal

11:10

but I don't think it was actually officially agreed upon

PyGamer0

wait why is Seggan making a vyxal implementation in java

lyxal

let me poll everyone

CMP: Scala for version 3? (Asking because it wasn't actually officially confirmed)

3

!!/hyperping

Vyxal Bot

@AaronMiller @AaroneousMiller @Allxy @Ausername @Milk @PyGamer0 @Wansen @astonearachnid @emanresuA @grandBagel @hyper-neutrino @math @pxeger @rues @user ^

lyxal

@PyGamer0 it's a side project

also, I need to be on the ping list

!!/ping me

Vyxal Bot

@lyxal I have put you on the ping list.

lyxal

11:11

thank you vyxal bot much appreciation

mathcat

wait I'll try something

PyGamer0

~~lets use C++/Rust for v4~~

mathcat

!!/ping me

Vyxal Bot

@mathcat I have put you on the ping list.

mathcat

aha

PyGamer0

11:12

!!/pong me

mathcat

I'm a twice now in the ping list?

lyxal

@mathcat no

PyGamer0

!!/ping me

Vyxal Bot

@PyGamer0 I have put you on the ping list.

lyxal

@mathcat well kind of

PyGamer0

11:13

!!/ping list

lyxal

@mathcat your old username is on there

which just happens to be a subset of your current username

mathcat

oh no

lyxal

meaning that it will still ping you

mathcat

go away math

lyxal

but you only get one notification regardless

mathcat

11:13

!!/don't ping me

Vyxal Bot

@mathcat I have taken you off the ping list.

lyxal

@mathcat only removes your old username if you currently have that username

math is still on the list

mathcat isn't

mathcat

@hyper-neutrino Help me defeat math!

PyGamer0

@mathcat did you mean "mathcat"

mathcat

no "math"

and math too because maths is better

Seggan

14:46

@PyGamer0 wait you're using Scala? Is it gonna be interpreted or compiled?

!!/ping me

Vyxal Bot

@Seggan I have put you on the ping list.

PyGamer0

15:03

@Seggan i am not using scala

vyxal v3 is planned to be in scala

because of @user

user

15:38

@Seggan interpreted

Seggan

mhm

user

I twisted lyxal’s arm :p

Scala compiles to JS so we can have a frontend-only interpreter on github pages

I’m also thinking of adding a debugger

Interpreting may make debugging easier

@Seggan is that a good mhm or a bad mhm?

Seggan

just a neutral mhm

I don't really care, I've just run speed tests of Jyxal and Vyxal, for the same program, Jyxal is on average ~130 ms, Vyxal ~2050 ms

*1050 ms

hyper-neutrino

!!/rm ping math

Vyxal Bot

@hyper-neutrino done

user

15:50

@Seggan Damn, nice

I suspect the Scala impl will end up being around 1000-2000 ms

Nowhere near Jyxal, though

Seggan

Tthe program was ⟨1|2|3⟩:+1+, if you're interested

Although including the time taken to compile increases Jyxal's time to ~540 ms

still 2x faster

user

That's still pretty good

mathcat

16:09

@hyper-neutrino thanks

!!/ping me

Vyxal Bot

@mathcat I have put you on the ping list.

Razetime

16:27

github.com/Vyxal/Vyxal/issues/689#issuecomment-1023807987

CMP: Relating to above link: is the K flag still wanted?

4

Seggan

Flag R: Treat numbers as ranges if ever used as an iterable. Isn't that already the case?

vyxal.pythonanywhere.com/#WyIiLCIiLCIxMChuLCkiLCIiLCIiXQ==

lyxal

22:58

@Seggan no, sometimes numbers are treated as a a list of digits

Transcript for