« first day (448 days earlier)      last day (1327 days later) » 

00:49
am i losing it?
do lambdas auto-eval?
Yes
It's a new feature
oh
why?
Only when printing
00:50
so that it doesn't print lambda:13269762369712396713879giasuigd8792q78etqyg v8hegt2?
01:02
Yep
so when does it not return 0?
does it return the last pushed value?
Correct
01:32
@Underslash actually that's a big with implicit input and lambdas
@lyxal a big problem?
Idea: Use scala or kotlin instead of Python. I see no way in which this could inconvenience people or set back work on vyxal
@Underslash it's not that big
im so confused lol
Don't be
Just know that lambdas don't exactly play nice with implicit input at the moment
01:39
ooh
thats what you mean
Correct
oh thats weird
i have: `?λ1|n;`and the input is 26, so it prints 26
but when I tack on a , it prints 0
Of course, it works with explicit input
@Underslash oh
Uh
That's not expected behaviour
yeah thats what I figured
 
2 hours later…
03:12
Do you think that's accurate?
Also, @pxeger, how long would you say the code clean up would take?
A day? A few days? A week?
Asking because a time estimate would help chose a starting day
For example, if it would only take a day or two, the halting of production could start now
But if it would take longer, we'd probably have to wait a week for the recent version to stabilise a bit
@lyxal timwi, creator of the IDE made that
@Razetime what?
The link goes to a comment on the vyxal community ad
hm it centered on the esoteric ide ad
ppain
 
2 hours later…
05:16
@hyper-neutrino @Razetime how's this for the community ad?
good
much better
The background is darker, the contrast it better, and it looks a tiny bit cleaner
05:36
@lyxal have you considered rewriting vyxal in C
or C++
why would you do that to yourself
Vyxal is staying in Python for the time being
time being = indefinitely
if I was to migrate to another language, it'd be a dynamically typed language
@hyper-neutrino why not
@lyxal then rewrite it in LUA
alternatively
no
hmm why doesnt anyone want to rewrite their projects in C
05:42
C is for masochists
> masochists
what is that
> A willingness or tendency to subject oneself to unpleasant or trying experiences.
there's a slightly more NSFW definition too
 
1 hour later…
06:50
@lyxal That looks really good, but what is the code actually supposed to do?
also, something is broke with the online interpreter
if you have code, then comment it out, if you implicitly output implicit input, it uses the commented out code's result instead
07:28
@Underslash that looks like an issue with the v2.5.x+ parser
Jul 4 at 23:54, by lyxal
Because the lambda is hyper, the 42 is AMiller42, there's a (was)if statement, kN is the (raze)time and username is you and A Username
If anyone has a better code snippet for the community ad, feel free to share it here
what exactly is the string snippet?
It's a compressed string
ooooooohhh
y'know
you would think I would know this by now lol
I was looking at other snippets, but in terms of syntax elements + the references, I think it works really well
λ42=[kN|`ʁβ`
were comments just not comments?
the parser wasn't looking for them properly
thats kinda funny
@Underslash what was happening was that if # was followed by a newline, it'd run forever.
because a) I had the condition wrong and b) I wasn't moving to the next character
that'll do it
generally moving from one character to another is good when parsing ;)
07:54
hmm
I know
especially when you're using a while loop + continue
while True:
pass
there's a v2.5.x+ related bug tag for a reason
@lyxal btw the cache problem I had is fixed now
@math good
@pxeger When do you suggest the cleanup start?
There was some poll
07:57
@lyxal user posted a strawpoll
first
so is the code cleanup just a refactoring of everything to make it not spaghetti?
@lyxal thanks
@Underslash it shouldn't be too spaghettified knowing lyxal
refactoring in the sense of making it maintainable by anyone
yeah thats def a good idea
what are people working on rn that would need to be finished before the cleanup could occur?
idk tbh
i'm not part of the cleanup side
pxeger wants some bytecode thing
and the main task looks like running black on the whole thing\
and also refactoring into multiple files
08:03
@pxeger do yall have any idea how its going to be split up right now?
@Razetime that's easy
I do, in my head...
maybe just decide right now
just use VS Code
@Razetime ^
python mode also uses black so /shrug
08:03
couldn't you also do something with github actions to automatically run it on it when pushing?
yes, I've put that on the project
I set up VS Code to autoformat with black on saving
y'all should do the same
BTW, what line length did you run Black with? The default (88)?
imagine not using the chad atom lol
@pxeger yeah, the default
@Underslash I used to
but then I switched away
08:05
im assuming for larger projects / where stuff had to compile?
no
what for then?
I just switched because Atom stopped working
and because VS code is actually cooler
@pxeger if I format with black, does that mean that flake8 will mostly be happy?
vs code is straight up better and effectively maintained by the same company now
@lyxal maybe, but how did it stop working?
08:06
@Underslash I don't know. Something about javascript
@lyxal it will complain about things that are bad practice but that can't be fixed automatically with tools like Black
e.g., using eval everywhere cough
what are some of the pros of using vs code versus atom?
@Underslash can Atom be set to auto-format your code upon saving it?
I'd imagine theres a package for that tbh
@pxeger what's the overall plan for the clean-up?
08:11
@lyxal is it normal to have a red no in the local version of flask?
Yes
yes
that is the george webring
That's the George part
Ninjad
which george?
@lyxal Initially I/someone would run all the autoformatters on the code and refactor it while noone else was working on it (to prevent merge conflicts), and then incrementally fix minor issues like adding type hints, fixing lintier complaints, improving comments/documentation
But you've already run black on everything lol so...
08:12
@pxeger but we're still pausing work though
Because the refactoring still needs doing
Black was the easy part. The comments and static types and documentation is the hard part
ok well I'm working on that now
@lyxal if we add those one or a few at a time in individual commits then merge conflicts should be manageable
why are the first 100 factorial pre-compiled?
does it really make that much of a difference
it's by far the fastest way to do it
Probably not
(because people probably don't use factorials that large)
But yes, it is more efficient
I mean, even with large factorials, they should take like microseconds to calculate
08:15
math.factorial takes like
0.031s real time for the first 100
@lyxal hopefully a couple hours
sympy takes about the same
but realistically about a day
@hyper-neutrino so does the first 1000
and then when you switch to 10000 it stalls
because they're likely cached in CPython
08:18
mhm
it also uses some fancy divide-and-conquer algorithm for computing them
doing the factorials manually take fractions of a second on idle on my machine using the most simple factorial approach
like, im not sure it really saves time vs just not having a large cached table there
anything below a 1000 is very fast to calculate, so having the first 100 makes little difference
@pxeger just make sure you only do as much as you want to do. And leave some stuff for me!
lol :)
@pxeger what's with html?
08:29
I don't know, what is??
@pxeger cleaning up. are you going to do that too?
oh
maybe
but then black can't do it
we'd use Prettier or something to do HTML
if I were in charge of Vyxal I'd ditch the online interpreter entirely and put it on ATO, tbh, so I'm not that bothered
@lyxal make sure you use isort --profile black to get them to agree
 
1 hour later…
09:38
In vyxal, where are all the constants defined?
except command.py
I see only functions in Vyxal.py
no, but there it's not actually defined, is it?
I thought it was for debugging
nope it's definitely defined there
quite obviosu why it's done that way
ok
it's parsed into an ast and then the transpilation is done by lookup into the commands
 
2 hours later…
11:34
@Ausername I may or may not have done a tiny bit of sandbox sniping
I mean, I got it working pretty quickly anyway, so it doesn't really matter ;p
@lyxal Oh
Btw 16
that's 18 though
that's the same link
Is it? Oops
Well 15
what an epic gamer
Beating osabie woohoo!
Dammit
11:50
@Ausername what.
how?
how are you so epic?
Magic
And the experience of ~150 Vyxal answers
Why user?
because we have a culture of rickrolling here
They really shouldn't've made it multiple-choice
yeah
you're probably right
Oh nice
@user Hi
11:56
@Ausername Why not?
I figured itd be better in case people want to vote on a range of days
@Ausername also, hi
@Ausername :)
12:50
@lyxal Why do the backticks look different from each other?
I honestly don't know why
Is this broken for anyone else?
@pxeger how do I set up the Vyxal package?
I tried pip install --upgrade --user .
but it didn't work
you don't need to do anything, you should just be able to do (from the root of the repo) import vyxal.interpreter
13:06
So then how do I run the repo?
*Repl
the Python repl?
no
the Vyxal repl
I didn't even know vyxal had a repl lol
previously, running Vyxal.py let you use it as a repl
but
python -m vyxal.interpreter
13:09
oh
very nice
thank you
btw I think Vyxal should move away from auto-installing a whole load of pip packages as soon as you start the interpreter
it didn't use to - it only did it if they weren't installed
even that's pretty narsty though
Is the online interpreter being really slow for anyone else?
@AaronMiller yes
not even the webhooks can connect
13:16
do you know why?
no
it just says "We couldn’t deliver this payload: failed to connect"
I think it's with pythonanywhere, because the Grok and brainbox interpreters are really slow too
yeah
I saw that pythonanywhere was taking a while to load
even pythonanywhere's homepage is very slow
13:17
^
when I first tried to get to the dashboard, it wouldn't connect
@AaronMiller but that was still a bug though
well, while PythonAnywhere is on fire, I think I'll go sleep
maybe when I wake up, it will have fixed itself
famous last words
o/
the l flag doesn't work on numbers
According to the PythonAnywhere twitter, it looks like they're getting DDoS'd at the moment
So that would explain the slowness
yoikes
13:23
bruh
 
1 hour later…
14:53
@AaronMiller No, what's happening to you?
Is everyone fine with starting the code cleanup roughly around July 16?
@user It looks like lyxal fixed it.
2 hours ago, by lyxal
@AaronMiller but that was still a bug though
is now a tag on the main site!
6
Nice!
@lyxal ^. feel free to edit the tag wiki to make it more descriptive if you want to add things about the language or whatever
@AaronMiller Oh good
15:04
@hyper-neutrino noice
Just submitted a tag wiki edit, someone please review it
both have been approved
:)
16:05
Idea: "^<>Vyxal", a 2D version of Vyxal.
V>XA<
looks very vaguely like Vyxal
and V is down, > is right, A is up, < is left, and X is all four directions
16:24
VYXAL seems kinda like it would fit with a 2d language
16:42
Oh wait, y'all started on the code cleanup without me already? :/
 
1 hour later…
17:43
Could I help with the comments btw?
I love comments
Sure, go ahead
Okay, so in the main branch?
Or is there a different branch for the cleanup?
 
1 hour later…
19:03
@AaronMiller What's a 2D language?
a language where the code forms a 2D structure (pretty much always just a grid) where the code being run depends on the character under the instruction pointer (IP) and the IP can move around that grid
@math Befunge is considered the first general-purpose 2D language, if you want to check it out. You can also check out Grok and brainbox, which are a couple of 2D languages I've made.
 
4 hours later…
23:30
@math if you're that much of a masochist then sure :p
@user yes
Aight great
All I've done so far is install the black plugin for VS Code
And now it automatically formats my code every time I save it
It's beautiful, isn't it?
Is the fizzbuzz issue in #140 fixed? Or does the pr need a bit of work before merging?
It's not, unfortunately, lemme work on it a bit
I have a feeling it's one of the imports messing something up

« first day (448 days earlier)      last day (1327 days later) »