« first day (4132 days earlier)      last day (44 days later) » 
00:00 - 23:0023:00 - 00:00

12:21 AM
@RadvylfPrograms may I suggest cookies?
Then you can donate them to Vyxal
@NewPosts I do hate zero-padding, and I do love finding clever ways around it
Hmm if only there were something that padded one string to the length of another with spaces and then something else that changed those spaces to 0s :p
I know about that, but it doesn't do list
Idea: Scratch *= Lisp
12:41 AM
oh yes
1:19 AM
@hyper-neutrino for the last 10mins, the voting fluctuated between being tied, 1 vote apart and 2 votes apart :P
Made it very annoying for the announcement post, as I had to be change the "dead tie" bit in case it wasnt
1 hour later…
2:27 AM
one of these is not the like other
one of these things doesn't belong
can you tell which thing is not like the others?
2:44 AM
Q: Generate All 8 Knight's Moves

Aaron ThomasIntroduction In chess, Knights can move two squares in one direction, and then one square in an orthogonal direction. On an x/y plane, these can be thought of as being a relative change of coordinates, such as (+1, -2). Challenge Your task is to output all 8 possible moves as pairs of relative x/...

3:03 AM
@lyxal there isn't a way to get reputation quick, duh :P
Precisely what I meant! 4 years and I still don't have the optimised rep grind :p
@lyxal I do, I just can't be bothered to use it :P
3:39 AM
Kinda sus
Very sus
in other news my school will start in 9 days
*sad PyGamer0 noises*
In other news my school will start in 17 hours
Y'all nerds going to school. Smh
@PyGamer0 you said you were in the codespaces beta. Have you had a chance to try it out yet?
3:50 AM
Hey, we don’t have a Daniel
@PyGamer0 lol I now know where this is from
4:15 AM
I’m in deep, implemented groups to the fork of a portable Regex library for c.
yikes I'm expected to actually do stuff in app inventor
“I just want it to do exactly what I want, is that so much to ask”
4:35 AM
@lyxal where :P
@lyxal no..
@PyGamer0 codespaces
@lyxal nah...
it was Protip: Great commit messages should never be more than 50 characters..........
4:52 AM
@lyxal i am going to use codespaces now..
(yes i installed the vim plugin)
5:12 AM
in canvas, 34 secs ago, by emanresu A
Anyone want to help me draw a giant donut?
2 hours later…
7:38 AM
i need help with regex
i want it to recognise the three lambdas separately..
i did it
import re
def transpile(x):
  x = re.sub(r"λ([a-zA-Z0-9_,=]+)\.", r"lambda \1:", x)
You can use [^λ] to specify “anything but lambda”
@ATaco i guess that is more cleaner...
and better
Also, you probably want to use non-greedy match instead
So λ(.+?)\.
Which will match everything up to the first dot
@ATaco i also have to replace in the group
so i have to take that group and replace in lambda \1:
7:49 AM
@ATaco ok that works thanks
@ATaco and i think i will change that + to *
Up to you
@DLosc say, how did you go about launching and spreading Pip? Transcript diving doesn't seem to have any introductory message, and it doesn't have a chat room either.
I can't seem to find anything about the history of it
wats pip
Not to be confused with the python package manager also called pip
@lyxal lmao precisely why i was confused :P
dang it was made 7 years ago
@lyxal It's quite difficult to search but I don't think it's widely used - DLosc and Razetime are all I can find
And I get stuff like:
A: Garden architecture - ASCII style

R. Kapdc, 210 197 bytes [256r^1-255/]sx?dddI/dsT9r-sYI%ddIr-sqdsesm-I/sN[[lNlxx124*PIlN-lxx35*PIPlq1-dsq0<o]dsoxlN1+sNledsq0<oq]sJ50!<J[Ilxx35*PIPlY1-dsY0<E]sElY0<E[lmlxx45*PIlm-lxx35*PIP]sClTI>C[Ilxx45*PIPlT1-dsT0<Z]dsZx Try it online!

I've seen other people using it from searching for it
So it's popular enough to qualify for studying for my blog post
yeah :P
500/800ish of the answers are DLosc's
Although most have 0/1 score
71 are Razetime's and 129 are from Add a Language To A Polyglot
other people trying it is probably at least in part because of the emphasis on true precedence-based infix, and it overall being golfy without reliably keeping up with the big dogs--or trying to
I more want to see if it somewhat follows my hypothesis of popularity/how it matches my theories
Because the mote data points the better :p
9:05 AM
in other news i might need to give dropout regularization a try for this nn i'm training because i can't help but feel like maybe i shouldn't have this many weights on the order of 10^-315
and maybe tune the size of each layer/number of layers while i'm at it :P
Maybe do some research into Jelly from back in the day
@UnrelatedString One day I wanted to learn about neural networks so I looked up a tutorial. "with your basic understanding of multivariable calculus" - Hm, yeah nah
you only need the calculus for training them but you don't actually have to understand the calculus to implement the backpropagation
you also don't have to implement the backpropagation :P
there are... a lot of libraries
@emanresuA I already am
What have you found?
I'd think it would be pretty revolutionary and kinda scary to those who were used to second-generation golflangs.
It matches my ideas of pre-launch, and fits some of the theories of post-launch
9:10 AM
and it sounds like it would be complicated to even describe mathematically, until you realize that running one layer is literally just multiplying a vector by a single matrix then applying a scalar function to each entry of the result
that second step sounds fancy when you call it an "activation function" like you have some physical neuron processing inputs but you're literally just throwing some nonlinearity in the mix so your whole net doesn't reduce to just one matrix multiplication
@lyxal ooh
@UnrelatedString I'd share an overview of what my theories on the best post-launch marketing practices are, but I don't know what they are yet :p
Those are for me to figure out later tonight lol
But as linked in the draft gist, they fall under the categories of initial engagement, expansion and maintenance
Initial engagement being the first weeks and months after launch, engagement being how new users are drawn in, and maintenance being how to ensure things don't fall apart once you've grown
And that's all I have that's concrete at this stage
9:42 AM
i'm looking forward to especially your conclusions about maintenance, because jelly's a very weird case with that
it's going strong despite not having updates in years, but in its early stages it was updated *heavily*--although it was competitive with the other golflangs of the era even when half of the most obvious builtins were missing, it kept getting refined from there
Jelly's a weird case because it was the start of its own generation
And I feel like we're on the verge of a new generation of golflangs with Radvylf's and pxeger's ideas
Speaking of which, @pxeger are you going to tell us what your golflang idea is? (radvylf has)
10:23 AM
if i ever get around to actually making perhaps i feel like it'll very solidly sit between generations
the idea is like 70% just making systematic improvements over jelly but then the remaining 30% might be a bit on the wonky side
10:43 AM
@UnrelatedString the way I see it, Jelly had what I'll be terming an external factor - influencers
It had the full force of Dennis' moderacy until 2019 and prominent users who are able to actively recruit others kept it going during 2020 and beyond
social media jelly influencers
10:54 AM
oh yeah having dennis's clout behind it definitely helped it grow lmao
active moderator and praclang golfer makes a strong golflang, what do you do, not try it
That's one of the reasons I was extremely happy when hyper joined vyxal lol
Mod affiliation is like winning the lottery and busting everyone else at a poker table lol
Both at the same time
very insightful, thanks
11:03 AM
@lyxal lol
@lyxal lol
Q: Can a run ascending list be made?

Wheat WizardA run ascending list is a list such that runs of consecutive equal elements are strictly increasing in length. For example [1,1,2,2,1,1,1] can be split into three runs [[1,1],[2,2],[1,1,1]] with lengths [2,2,3], since two runs are the same length this is not a run ascending list. Similarly [2,2...

@UnrelatedString forgive me for this but is that /srs or /j
11:18 AM
> <div></div>
Oh wait you meant that
yea h lmao
Lol I thought you were saying that about my observations about jelly lol
That's why I couldn't read the tone lmao
11:23 AM
Gotta love adding lmao to things that have no humour at all and somehow making it funny lmao
@UnrelatedString took me a while to figure out but now i see that it means “everything with a start has an end”. Truly insightful
@lyxal there were jelly tiktokers? :o
i hanged my terminal with PW halp
i just closed the terminal..
@PyGamer0 please don’t do this
This mangles strings containing lambda
@user umm i realised that very late..
11:30 AM
Regex is not the tool for parsing complicated stuff

☺=lambda x: lambda y: x
☺=lambda x: lambda y: y

☺=lambda p: lambda q: p(q)(p)
☺=lambda p: lambda q: p(p)(q)
☺=lambda p: lambda x: lambda y: p(y)(x)
☺=lambda p: lambda q: ________not(p(q)(p))

def fn2bool(b):
    return b("True")("False")

f________or i in [true, false]:
    f________or j in [true, false]:
@user you should have warned me earlier ^ :P
There are good libraries out there to help you
Oh god what did you do
oh no
What rule inserts those underscores?
11:31 AM
and this is why you don't use regex for that. use a library which covers up the regex instead!
@user several rules individually
@user just a few replaces
before lambda, and, or, and not
A \b can fix that
But this is still incredibly cursed
whats a \b ?
11:33 AM
@user ^
just because you can fix it, doesn't mean you should
Word boundary
@UnrelatedString i mean the whole thing can’t be fixed
Only tiny parts
@NewPosts i need more brain cells to solve this
Unless you use an insanely powerful regex flavor
But then you’ll probably go crazy even if you succeed
@PyGamer0 yeah it's deceptively hard to do without going max naive brute force
11:34 AM
@PyGamer0 you a zombie?
@user no i was diagnosed with Zwei-smadzenes-cellules :P
(guess what that means :P)
@UnrelatedString thing is i cant figure out how to brute force it ...
i mean you can just generate a list described by the given element counts and then test every single permutation for ascending run lengths
@UnrelatedString yeah i was thinking about that...
i thought i might be on to something until i noticed that the run lengths don't have to be consecutive lmao
actually maybe you could do something close to the intuitive explanations in the spec
like you could specifically split numbers around other numbers
11:53 AM
does jelly have a builtin to check if the input is sorted?
you can do it in two bytes :P
If only there were a quick that returned whether or not something stayed the same after applying an atom to it
well i forogt that exists
that sure would be useful in a variety of circumstances due to its interactions with vectorization and ability to be used with other quicks among other things
yeah, that sure would be useful
but that'd be asking too much of Dennis
11:56 AM
reminds me just want to make sure all of y'all saw ߀NÐeSµ)Ƒ¡ on the deep alternating sums challenge
part of me is sad that it works
> Bendesufi
Sounds like a cheap knock-off of emanresuA
@UnrelatedString i think S⁼#'Ġ¥ᵟᶠW brute forces it giving all possibilities
(you can obviously guess the language)
if name == "Ben"
    echo "Benです";
i don't think that's valid in any actual shell but you get the idea
@PyGamer0 pog
idk if it works lol
@UnrelatedString Benare?
is Ben?
12:01 PM
try not translating it
also the highlight is the fucking µ)Ƒ
like that is just not how i'm used to thinking about )
@UnrelatedString you're gonna need to explain the joke to me then
my humour detector seems to be broken tonight
> Ben desu
anyways i am way the fuck too sleepy to actually implement the number-splitting idea for the run lengths challenge so uh yeah
Hey guis
@mathcat hello
12:05 PM
I'm getting better at nerdle
it's far more interesting than wordle
@mathcat that means you're a nerd.
you're welcome
I'm actually going to post a challenge now
that's what I call gaming
12:07 PM
@UnrelatedString explanation:

        W  ⍝ N copies of sequential indices of elements
      ᵟᶠ   ⍝ filter keep over permutations
     ¥     ⍝ last 3 links as monad
    Ġ      ⍝ group adjacent elements which are equal
  #'       ⍝ length of each
S⁼         ⍝ is sorted?
(if i know chaining properly)
someone doesn't like lyxal as much as they like pxeger and emanresu
Oh wait, new room owners!
I'm italic now :p
congratulations, emanresu A, lyxal and pxeger!
@lyxal epic
@lyxal no, you're lyxal
12:26 PM
now for the hard part of my blog post - the post-launch era
12:45 PM
@lyxal how is it hard? you are the creator of vyxal which has 100 stars on github
yes, but do I look like I know what worked and what didn't?
the pre-launch is easy to manage. it's post-launch that is touch-and-go
@lyxal ... yes ... ?
there have been things that have happened without interference on my part
i mean i guess the vyxal bounty drew some people into vyxal ... ?
e.g. vyxal somehow getting the 12 votes it needed to get LoTM this time last year in just 2 weeks
and there's been parts that were not my doing at all
12:49 PM
@lyxal like?
a) moderator affiliation, b) Chris Pressey (creator of befunge) starring vyxal, following me on github, c) expansion into code review, d) obtaining 100 stars
a) hyper joins vyxal
@lyxal woah
b) Chris Pressey's followers see vyxal
c) .... i have no words ....
d) very cool
12:56 PM
I can explain about half of those
these users are from a rickroll language I starred once
this user seems to have found one of us and followed through with the rest of vyxal
and the last one is me
also cookie.

Very funny joke.

50 mins ago, 48 minutes total – 3 messages, 2 users, 0 stars

Bookmarked 12 secs ago by PyGamer0

1:00 PM
what about it?
what is that
what was supposed to be a python like language made by some people from a discord (truttle) server
1 hour later…
2:19 PM
hey wow they made a decent stack status page
stackstatus.net is finally useful
but of course still no news on any quarterly community roadmaps for 2022
3:11 PM
@lyxal I basically just started using it. My first mention of it in chat is from five months later (and immediately garnered a "Why'd you name it the same as Python's package manager" comment... sigh).
Got it on TIO about a year after first release.
I've never done much advertising, which is probably why Pip has never become popular (that, and the fact that it rarely beats other golflangs). I'm hoping that LYAL and eventually LotM will help. I think the people who have tried it have liked it.
3:26 PM
Aug 26, 2016 at 20:16, by Rohan Jhunjhunwala
Im so happy for the new room owners
yes but currency ¥€$
4:02 PM
That's interesting: the messages in chat about new ROs have 13 stars each, but the actual meta announcement only has 10 votes
(of course, as soon as I say that, it's score becomes 11 :P)
I didn't know there was a meta announcement
I figured the link was to the voting
Ah, that makes sense
Nah, I followed tradition from the last 2 RO elections and posted an announcement, just to make it 100% clear who ended up being appointed
1 hour later…
5:15 PM
Today on Why Is JavaScript?: destructuring in rest arguments! ato.pxeger.com/…
What's the code?
ATO's blocked for me
function foo(...{2: third, ...rest}) {
  console.log({ third })
  console.log({ rest })
foo('a', 'b', 'c', 'd')
Semi-reasonable if you think through what it's doing though
it kinda makes sense as a generalisation of both ... and destructuring features, but their intersection is still weird
It does have its use-cases though: ato.pxeger.com/…
// typescript
function foo1(x: number, y: boolean, z: string[]) {
  // pretend this does something interesting
  console.log({ x, y, z })
function foo2(...[firstArg, ...rest]: Parameters<typeof foo1>) {
  return foo1(firstArg + 1, ...rest)
foo2(3, true, ['hello', 'world'])
an easy way to do a type-safe wrapper function which modifies some specific arguments
Non-sequitur: My congratulations to @lyxal for becoming an RO!!
5:30 PM
and not the other two of us?
No, all of you of course.
@lyxal is awesome because they invented Vyxal though, and Vyxal is my fav lang.
So first to mind.
5:43 PM
tfw replit decides that your repl is experiencing a DDOS even though it's not
@Ginger Do you never program on your own computer?
I often can't because reasons
plus replit does free site hosting :p
5:59 PM
@Ginger School/work computer sort of reasons, or potat comput reasons?
@pxeger Doesn't this kinda defeat the point of static typing?
It's far less readable and more likely to contain bugs than the non-statically-typed way to do this
(Assuming TS lets you tell it to ignore the typing when it's convenient, which I'd hope it does)
@RadvylfPrograms school/work
Ah, rip
(promise me you won't use that to doxx me lol)
Breaking news: Ginger goes to a school and/or works somewhere!
1 hour later…
7:10 PM
Should Leaky Nun still be an RO after their extended absence? IDK much about RO protocol.
They're not an RO
You might be looking at "frequently in room"
Got it. Yeah; the chattiquete link confused me.
Foolish of me.
Chat's UI isn't the best, not your fault :p
And it's not like "frequently in room" is a particularly memorable festure of it either
7:30 PM
List of absurd GitHub issue titles #6627:
> Fixed a bug that made - do addition instead of subtraction
8:11 PM
@RadvylfPrograms Not at all; it means you can change foo1's parameter types and foo2 doesn't need changing at all, but the call site still has to change
If you did foo2(x: number, y, boolean, z: string[]), then if foo1 changes, you'd have to change both foo2 and its call site
it's much more convenient, but you still get safety when calling the function
Hmm, I guess. It seems a bit to "look how cool this is" at the cost of readability, although tbf I do that all the time
Typescript's type system is honestly pretty great
If they fleshed interfaces out into full Haskell type classes (with derivability and default implementations), it would be close to a my idea of a perfect type system
8:27 PM
what do we think of a challenge about detecting various number classes using only regex
i know you can do, say, primes, evens, multiples of five, but i wonder if theres a way to make it interesting
ive got one rough idea but its not like, great,
something like "it has to match strings that are the class of the number followed by the number it is, and youre scored on both shortness and how many number classes you can do"
its got two metrics so its bad
but the concept intriques me :Pc
CMC: Is it a fibonacci number?
@RadvylfPrograms Pip, 14 bytes: Wi<ai::o+:ii=a
8:54 PM
@RadvylfPrograms Vyxal, 3 bytes
Honestly I have no idea how it terminates for non-fibonacci
That's "is it in the infinite list of Fibonacci numbers" and I don't know how it determines that something isn't.
any chance the search knows it's ascending
I expect Contains with a lazy list assumes the list is ordered, so it keeps looking until the elements get bigger than the value it's looking for and then it returns false.
oh yeah it could also assume that lmao
Try negating every other element and see if it breaks :P
9:08 PM
That approach in Haskell, 54 bytes
36 bytes (can be 35 if True & False can be switched): f 0 1;f a b n|a<n=f b(a+b)n|1>0=a==n
34 with infix: 0#1;(a#b)n|a<n=(b#(a+b))n|1>0=a==n
9:36 PM
JS, 41 bytes: x=>(s=v=>(v**.5)**2==v)(g=5*x*x-4)|s(g+8)
9:55 PM
Hmm, that's unintuitive. x is a Fibonacci number iff one (or both) of 5x^2-4 or 5x^2+4 is a perfect square?
Yeah :P
10:11 PM
CMQ: How many times have you been shocked by mains voltage (120/240)?
Well, probably more concerning in 240-land. Three times here.
Once when I was like 6 and putting up christmas lights, and then twice recently, which were sort of..."second hand shocks" from being too close to my idiot dad while he tries to fix outlets with the breaker still on
lol nice
10:15 PM
The first time was the only "real" sock I've gotten, where it feels like you're getting poked by needles all over. It doesn't really hurt, but it's a bit...shocking (no pun intended)
> sock
I remember touching the metal prongs of the television's power thingy as I was taking it out of the extension cord thing and it hurt but also felt kinda nice and tingly at the same time?
y'all are fucking wild
Sensible plug design and parents here
One of the times it was pretty minor since it was a GFCI outlet, and we'd gotten lucky since it was the load side. So I just got a metallic taste in my mouth. If it had been the line side things would have been more interesting.
10:16 PM
As Radvylf said, it's less concerning in the US
Yeah, but a lot more common since our plugs are garbage
for a split second i read that as ghci
I've luckily managed to avoid that sort of shock since I'm really careful when unplugging stuff, but I have witnessed some...interesting things
I touched an electric fence, but that’s about it
Oh this just reminded me, I went to this science park (?) once and there was this thing with lit up handprints on it. Being a stupid kid, I went up and immediately pressed my hands to it very hard. It hurt like hell and I fell from the shock (pun intended). Got up and read that the exhibit was supposed to give you an electric shock
10:19 PM
I don't know why they even had it there
to give you an electric shock
@ATaco Whoa, did it hurt a lot?
@UnrelatedString Oh that solves the mystery, thanks
What really scares me are our dryer plugs though...just as unsafe as the normal ones, except scaled up by two, and often even harder to get a grip on (and also 240v and up to 50 amps)
10:21 PM
Has a been done where you can only insert/append/prepend characters into the previous answer?
@user wasn’t actually very painful at all, but it sent me hurling backwards
@emanresuA What would the goal be?
Idk, just an idea
Could be interesting but I don't think I've ever seen one
Probably have to enforce a language restriction as well, such as enforcing a different language each answer
10:24 PM
yeah :P
That or it’s all going to be the same language, most likely
@ATaco Or you could enforce one single language (perhaps made up for the challenge) for all the answers
that'd get too trivial :P
“Write the previous answer’s language name”
@user Once someone figures out how to sneak a comment in it's game over
10:26 PM
Unless the language doesn’t have comments
I really want more challenges with languages specifically made for them
You could totally make your own language without any comments
At the same time you'd want to avoid 1, 1+1, 1+1+1 etc
I want more challenges that ask you to make a language
^ too
Like the “make a seemingly unusable language” c&r
CMC: Given two points, calculate the M and C values of the line between them, assuming P1 and P2 is not perfectly vertical
10:32 PM
Ooh, idea: add a language to a polyglot but the language you add is hidden
C&R answer chaining..?
In other words, pretty much this:
Q: Overlapping Polyglots

Esolanging FruitUnlike many C&R challenges, this does not require a separate robbers post; the goal of the robbers is to crack the previous answer and then post a new one as a cop. As answerers, you will write a series of polyglots that look like this (each column is a language, and each entry is the output of ...

Well then do I have the challenge for you
But I think having all the languages as opposed to just 3 will make for a very different one
Opinions on whether it'd be a dupe?
10:51 PM
Q: Divisible subset sums

caird coinheringaahingInspired by the recent 3Blue1Brown video Consider, for some positive integer \$n\$, the set \$\{1, 2, ..., n\}\$ and its subsets. For example, for \$n = 3\$, we have $$\emptyset, \{1\}, \{2\}, \{3\}, \{1,2\}, \{1,3\}, \{2,3\}, \{1,2,3\}$$ If we take the sum of these subsets, we can then ask ourse...

@NewPosts Does the mathjax on this render properly for y'all?
Does for me
Works for me too
00:00 - 23:0023:00 - 00:00

« first day (4132 days earlier)      last day (44 days later) »