Vyxal

user

00:08

@VyxalBot What's wrong with var btw? @Seggan

AviFS

@Seggan Nice! Would this play nicely with the tabl interface?

Eg. Is it going to have an option to handle invalid programs and infinite output?

@AaroneousMiller That'd be awesome. For the time being, I'm putting Vyxal aside and gonna work on supporting another lang, and then take a break. But if someone wants to work with me to get Vyxal running, I'm down!

It lives here: github.com/AviFS/tabl2/blob/apl/exec/vyx.py

Seggan

00:24

@user i like my variables nice and explicitly typed so i can scroll up and immediately see the type of my var

@AviFS js has a precision of ~53 bits

user

@Seggan I'd say it's pretty clear what the type of jshell.get() is but alright

Seggan

i agree, but i generally frown down upon var

take for example var x = someObj.getSomeValue().getSomethingElse(); What on earth is the type?

user

I agree that var is not helpful when the type isn't clear from either the name or the expression, but whenever that's not the case, I like to use var

I personally find it a lot easier to read code with tiny types like var and int than Map<String, List<String>> even if it's more explicit

But that's probably because I'm used to Scala, and entirely a matter of personal taste

Seggan

¯_(ツ)_/¯ ig, but this is chiefly my project so ill be the one reading the code most of the time

user

Ye

lyxal

00:42

CMP: Do we want to borrow one of Radvylf's droplets for running the backend of the Vyxal online interpreter? It'd act as a sort of API and github integration on push to production would be enabled. We'd keep using Pythonanywhere for the front end.

!!/hyperping

Vyxal Bot

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

lyxal

(Also keep in mind that version 3 will be coming soon and that it'll be hosted not on pythonanywhere and probably on github pages if scala -> js works)

emanresu A

¯\_(ツ)_/¯ Pythonanywhere's not that bad

But if radvylf's willing, I guess?

lyxal

Using a droplet would mean that you could actually stop a program mid-execution

but using a droplet would mean reliance on radvylf, meaning vyxal wouldn't be fully independent anymore

with pythonanywhere, access to server internals is decentralised

user

I mean, if Redwolf ever decides we can't use his droplet anymore, we can always just go back to PA

I doubt he'd do anything malicious or whatever, and he's going to give you an account, isn't he?

lyxal

00:47

I suppose this brings up another question: with version 3 coming, what should we do with the version 2 sites? Keep them running or just let them drop off and let sites like ATO host v2?

emanresu A

Keep them running

We don't want thousands of broken links

And all it requires is logging in once every three months

user

Keep them running ^

lyxal

Well I'm open to switching to the droplet if y'all think that'd be best.

I'm not the best with major decisions so I'll wait for more opinions before confirming that we're using it

user

Hesitation is a pretty good thing when making big decisions

Seggan

^^

user

03:13

Ugh, IntelliJ

I'll fix it tomorrow

A standardized format would help

Seggan

It's 10 PM here, I've got driving lessons tomorrow, will add one at my next convenience.

Vyxal Bot

Seggan approved on PR #3 (Vyxal/Jyxal): "Any way to tack this on releases?"

Vyxal Bot

03:29

cgccuser merged PR #3 (Vyxal/Jyxal) (Vyxal:archive-jar → Vyxal:master): Create jar artifact in Gradle workflow

user

@Seggan Awesome, thanks

I like having an autoformatter I can run any time I like

Also, in a few minutes, the Github pages thing for VyxalS should be updated, so you may be able to run some (very) basic stuff there (addition, multiplication, calling functions, and not much else)

Oh crap, my parser's borked

emanresu A

04:16

Unbork it

emanresu A

06:04

FR: Apply to all permutations then flatten by one layer modifier

AviFS

07:00

@Seggan I don't quite understand the response.

emanresu A

09:00

Where did we put the codepage file again?

!!/info

Vyxal Bot

emanresu A

09:33

Can we make the default range for ℅ on a number just python's regular range?

It's just so much more useful.

Vyxal Bot

09:51

chunkybanana opened issue #757 in Vyxal/Vyxal: Merge `Þ*` into `Π`

chunkybanana opened issue #758 in Vyxal/Vyxal: Element to push a[:n] and a[n:]

PyGamer0

@emanresuA so random.randrange?

emanresu A

Pretty much yeah

It's just so you can easily do e.g. a 1 in 10 chance

PyGamer0

does vyxal have a range which is 2..x

so 4 → [2, 3]

emanresu A

You can do ɽḢ

!!/run 4ɽḢ

Vyxal Bot

@emanresuA ⟨ 2 | 3 ⟩

PyGamer0

10:02

hm

mathcat

12:29

@VyxalBot in the tests, it says ['abcde', 2] : 'cde', but shouldn't it output ab and cde?

lyxal

12:42

the test cases only check the top of the stack

mathcat

oh, ok

user

13:36

@VyxalBot @lyxal this will loop infinitely for lazylists

I don’t think they should be merged honestly

Vyxal Bot

cgccuser approved on PR #760 (Vyxal/Vyxal): "That test case makes me uncomfortable but alright. Useful element"

user

I feel like we should get rid of P* entirely because how often do you even reduce by cartesian product

Seggan

@AviFS tl;dr js has plenty of number "space" to avoid the Y2K38 problem`

Vyxal Bot

14:13

Seggan deleted branch Jyxal/archive-jar

Seggan

18:46

Ya think we need a NOP flag that is just a filler?

Transcript for