@DLosc Python has default args, so I don't think you'll need to special-case anything if you default the second arg to infinity (but then I guess you won't use the single-arg variety anyway?)

@hyper-neutrino Make it camel case and you've got yourself a proper Java name

not enough design patterns

inclusiveRangeIfRightArgIsNotInfinityAndRangeToInfinityOtherwise

@Bubbler Right, that's my point: I'm always calling it with two args.

How about from_to_or_inf

@user ...Factory :P

or simply from_to, which reads like "from a to b" and "from a to infinity"

class InclusiveRangeIfRightArgIsNotInfinityAndRangeToInfinityOtherwiseBuilder {
  InclusiveRangeIfRightArgIsNotInfinityAndRangeToInfinityOtherwiseBuilderWithLeftArg withLeftArgThatIsNotInfinity(LeftArg arg) {...}
}

class InclusiveRangeIfRightArgIsNotInfinityAndRangeToInfinityOtherwiseBuilderWithLeftArg {
  InclusiveRangeIfRightArgIsNotInfinityAndRangeToInfinityOtherwise withRightArgThatMayOrMayNotBeInfinity(RightArg arg) {...}
}

Or go for some metaprogramming and support writing from_(a).to(b)

@Bubbler Yeah, the or_inf part would be unneeded IMO. I was hoping for a name that had range in it somewhere, but maybe there isn't a good way to do that.

@Bubbler Ah, now there's an approach! X^D

@user AUGH MY EYES

If you really want the word range, you can do range_from_to I guess

@DLosc That's not totally accurate - a real program would have overloading and some BuilderFactories and more :P

Or maybe an infix method foo range_to bar?

@user A what now? ^_^

@user Scala overdose?

I guess if it's Python, it'd have to be foo.range_to(bar)

@Bubbler Impossible, I can never get enough Scala :P

Deciding between range_from_to and just incl_range.

Not sure if we can patch int to do that, but I think this should do the job:

class From(object):
    def __init__(self, n): self.n = n
    def to(self, m): return my_func(self.n, m)
From(3).to(inf)
From(3).to(5)

I'm like 98% positive that code breaks several Python style guidelines. :D

:P

I do wish Python handled infinity better. I mean, it's there, it's a valid floating point value, it compares properly... but the only way to generate it (AFAIK) is float('inf'). Any attempt to do 1/0, even if it's 1.0/0.0, just raises an error.

Aww, you can't add extension methods to ints in Python

@user Nope. :(

@DLosc A golfy infinity is 1e999

@NewPosts is it sufficient to check if the graph has no odd-length cycles?

or is that not how this works

I get why adding stuff to built-in classes is disallowed, but still, it would be really cool if you could.

@Bubbler Huh, TIL

@hyper-neutrino I think so

Though intuitively, it seems like in this particular case, it's sufficient to check for 3-cycles. I don't have a proof yet, but I'm not sure how a greater-than-3 odd cycle could be possible without a 3-cycle also existing.

Hint: I have a hint to give out to you guys, but I'll keep it secret until someone finds out :P

@DLosc good point

CMC: Given a list of integers, count how many repeated elements there are. For example, [1,2,2,3,3,3] -> 3 (one 2, two 3s)

I have 4 bytes in Jelly, but I think 3 is possible (haven't tested the one I have in mind)

@Bubbler TIL Infinity=1e999 :P

APL, 5: +/1-≠

Nub-sieve, flip, sum?

@cairdcoinheringaahing Scala: x=>x.size-x.toSet.size

@cairdcoinheringaahing Same as your Jelly solution?

The 4 byte one, yes

@cairdcoinheringaahing Yes

Can the integers be negative?

(Well, I have 2 4 byters. One counts zeros, the other is flip and sum)

or zero?

@cairdcoinheringaahing Why is it 4 and not 3 bytes? I thought since they were all monads, it'd just be a 1-1-1 chain

@user Nub sieve is two bytes

@Bubbler why?

Oh, that's too bad

@cairdcoinheringaahing Pip, 7 bytes: #g-#UQg

@cairdcoinheringaahing Because this

Yeah, ŒQCS works for 4

@Bubbler Ooh, that's clever

Because I like general solutions, I'll say that yes the input will be any integer, but brownie points for that

@Bubbler What does that do?

@cairdcoinheringaahing One day, everyone's going to redeem all the brownie points you gave them, and you're going to have to do a lot of baking :P

I have ĠẈ’S

@DLosc "Return a Boolean array with 1s at the indices in z" and sum it

@user That's fine, I like baking :P

If we ever have a CGCC user meet up, I'll bring brownies

œ-QL also for 4

Yeah, so many fours

@Bubbler Oh, so that's "count unique elements," not "count repeated elements."

Aww, my suspected 3 byter doesn't work: Try it online!

@cairdcoinheringaahing I personally like my brownies served raw, not baked >:)

not sure why Lƙ’S doesn't

NGL, I refuse to touch ƙ with a 10 foot pole

same

What's so bad about it?

@user If you don't kill the fae, they'll steal your soul

@user miles coded it, didn't explain it to anyone then proceeded to use it in crazy ways to save bytes

lol

@cairdcoinheringaahing Aye, and replace your young with changelings

Luckily, I have no such weaknesses

There's 2 ways I can figure out what a builtin does in Jelly: see it in action or read the source code. ƙ never makes sense when used in code (I mean that, it's a dyadic monad), and the source code is stupid dense

IIRC Dennis doesn't even know what it does

There's a function in the Jelly source code (split_key I think) that is only used for this builtin, except the name doesn't fit with the quick description, and I have tried to read the source code to understand what it does and failed

Chat, I hate you

Uh, so the two byter was wrong anyway

def split_key(control, data):
	groups = {}
	order = []
	count = 0
	for key, item in zip(control, data):
		key = repr(key) if type(key) == list else key
		if key not in groups:
			order.append(key)
			groups[key] = []
		groups[key].append(item)
		count += 1
	result = [groups[key] for key in order]
	if count < len(data):
		result.append(data[count:])
	return result

@Bubbler Yeah, DLosc pointed that out :(

	'ƙ': attrdict(
		condition = lambda links: links and links[0].arity,
		quicklink = lambda links, outmost_links, index: [attrdict(
			arity = 2,
			call = lambda x, y: [monadic_link(links[0], g) for g in split_key(iterable(x, make_digits = True), iterable(y, make_digits = True))]
		)]
	),

That's the source code for ƙ, if anyone can decipher that, hats off to you

quad diamond is confusing me :c

@hyper-neutrino Starting with ({⊂⍵}⌺2 2)something should help you out

@hyper-neutrino Ah, your mistake there was using APL :P

(with ]boxing on)

Stick to a respectable tacit language like Jelly :P

@Bubbler i used that to visualize it but then when i try to use ({∪⍵}⌺2 2) i don't exactly get what it's even doing :p

Can't be dealing with a language that doesn't even have an "is prime?" builtin :P

@hyper-neutrino ⍵ there is a 2x2 matrix. ∪ on a matrix gives you unique rows, not unique elements, so you need ∪,⍵ for the challenge

@Bubbler that then gives four matrices so i'm still confused how it's working

@hyper-neutrino Code?

@cairdcoinheringaahing Okay, so split_key takes a list of keys and a list of values. It pairs each key with a value. Then for each key, it finds all the values associated with that key, puts them in a list, and returns a list of those lists.

({∪,⍵}⌺2 2)

And what's the input?

@hyper-neutrino What did you do with this review task? It says "Reviewed" but I don't see any activity by you?

upvotes

@Bubbler oh - full code is ({∪,⍵}⌺2 2) 5 5 ⍴ 0 0 0 1 2 0 3 3 2 2 4 5 3 2 6 4 4 7 6 6 4 4 7 7 7

Sandbox posts last active a week ago: Largest Number with No Repeating Digit Pairs, Notes, (untitled), COBOL(lite) interpreter.

@DLosc But then how does ƙ work from that? It looks like it maps the provided monad over each list but tio.run/##y0rNyan8///QzmMz////H22oY6RjrGOiYxoLYYNxLAA

Wait, is it "map monad f over the groups formed by grouping y by x"?

wait, k-hook always produces a dyad? o.O

yep

Yeah, I think you've got the lists backwards. TIO

Because tio.run/##y0rNyan8///QzmMzHf7//x9tqGOkY6xjomMaC2GDcSwA

i feel like this conversation happens every month or two but we're more shocked each time

That makes me even more annoyed tbh

wait that's what that is

i thought it ran on the lists of indices ._.

Lol

lol

you mean it isn't supposed to just map over consecutive groups?

Jelly needs a "group by" quick, but that got implemented as <link>¹ƙ$ instead of having a proper quick

yep

@lyxal Try it online!

@cairdcoinheringaahing w...what?

In which case, miles sucks at writing docs:

@lyxal The groups don't have to be consecutive.

it maps over groups, period

> Key. Map a link over the groups formed by identical items.

except they're thr groups of the uh

which argument

The fuck?

i just copy pasted the thing explaining them

and i already forgot

@UnrelatedString It's own values

in this case it's the same because left and right are the same

Basically, when provided one list, each element maps to itself as a key. Therefore, it's equivalent to just grouping the equal elements

yeah

@hyper-neutrino ({⊂∪,⍵}⌺2 2) 5 5 ⍴ 0 0 0 1 2 0 3 3 2 2 4 5 3 2 6 4 4 7 6 6 4 4 7 7 7 for better visualization

@UnrelatedString But it doesnt

oh yeah that quote is wrong in some way isn't it

@Bubbler oh ._. lol thanks

The quick doesn't do anything to the monad. It groups the left based on the right, then maps the monad over them

^

No, that's the same thing the explanation says.

doesn't it group the right by the left?

Honestly, if ¹ƙ was just implemented as an atom, that'd make more sense

yeah

¹ƙ` specifically, or am I misunderstanding

wait yeah i'm misunderstanding

It's not like Jelly is particularly short of unassigned dyadic atom characters

yeah, from my uh... now top voted answer *sigh*

:pensive:

@hyper-neutrino I love that you didn't even know how it worked :P

@UnrelatedString So it's similar to x f⌸ y in APL, but with less information (in that the key isn't passed into f)

Despite that being documented UB that we've discussed multiple times in JHT before :P

@cairdcoinheringaahing yeah, I was writing up a whole explanation for the entire link and how it coincidentally worked out and then I got your message that "it's just cuz u breaks everything lol"

looks like the last time we talked about this we determined it's based on a j adverb, so that j adverb was probably lifted from that

@UnrelatedString miles full on said he copied the J adverb

lol i'm dumb, I was using ≡ instead of ≢ like i meant to ._.

@hyper-neutrino Interestingly, it doesn't work in my fork :P

;€ is one of the most common digraphs in Jelly answers, so I assign that to u (unify), so the chain is broken as the h in hurt, and the y then errors

ah :p

@Bubbler got -4 bytes though you could probably do better; thanks for the tips and help :D

@cairdcoinheringaahing I'll give you the count of unique elements AND the count of repeated elements, in four bytes: Try it online! ;)

@DLosc Sorcery! Burn the witch :P

If ¹ƙ was (e.g.) h, then that'd be 5 bytes (assuming the quick didn't exist), solely because of chaining :/

Anyway, to circle back to something said a while ago: I agree that split_key is not a descriptive name. Maybe group_by_keys would be better.

CMQ: I'm making a regex-based language. Should I number the capture groups starting at 1 (like most regex flavors) or at 0 (taking advantage of an extra number that's unused)? Note that I don't think this language will need the ability to refer to the full match, which is what other regex flavors use capture group 0 for.

the primary reason i typically like 1-indexing in golfing languages is so that "not found" is falsy - not sure though if this is a golfing language or not, and also how much sense either would make in a primarily regex lang

Yeah, there's no concept of found/not-found; these are basically just backreferences like in the regex (.)\1* that finds runs of repeated characters.

Its golfiness is rather like Retina's: not exactly optimized for golf, but the syntax will probably be fairly terse because regex syntax is fairly terse.

hm.. I still feel like being able to refer to the whole match will be somewhat helpful maybe so my personal recommendation would be to just keep it at 1-indexing and use 0 for the full group. but I don't know much about regex and specifically using it for golfing anyway ¯\_(ツ)_/¯

The reason why referring to the whole match isn't likely to be useful is that this is a regex-match generation language (see my CMC from yesterday). You can't refer back to the full match while you're still generating the match. Now, it might become useful if I add some multi-stage functionality in the future... not sure.

I guess I'll leave it at 1, 2, 3. Extra golfiness if there are more than 9 groups isn't worth the surprise of having groups numbered differently from other flavors.

Honestly, we could do with a reflex based language that is heavily optimised towards golf

*regex

you're having some fun spelling that i see :p

I hate mobile

@cairdcoinheringaahing I'm working on some ideas there too, although it's not specifically regex based but more like a golfier version of V.

It'd probably involve writing your own regex engine to properly handle a lot of it, so it'd be a lot of work

well you could start with things like using a full codepage and overloading some characters that you probably won't need otherwise

like \d could be one byte :p just as a basic example

It is true that I don't really understand the arcane powers of .NET regex, so a regex-based language invented by me would likely not be as good as one invented by someone like Martin Ender.

Hello everyone

o/

CMC: Find the "reverse regex" of (...)(\1|(...))\3 without using (?!...), or prove that it is impossible. A "reverse regex" Y of a regex X is the one that satisfies "a string x matches X iff the reverse of x matches Y". A match here is a full match (^...$ is implied).

Isn't that regex equivalent to .*(...)\1, or am I missing something?

Oh... Is the ... supposed to represent something specific?

Scrap it, didn't realize it can backtrack

I know we have a golf for programs you can only run once, but is it a bad idea to suggest one that can run only twice? Then you have to find a way of determining if it's been ran before, instead of just a program that deletes itself

that sounds interesting for me. It's definitely different.

@SlamJammington there are a lot of ways of determining

Which would not often be very interesting

Kekw, just heard an ad on the radio for a "better" WiFi modem, and that using the faster internet would lead to "no more blue screens of death".

Like, even without the internet my stuff still gonna crash horribly. Stupid idiots in advertising.

lol tf

Netgear ad

How can the temprance movement prevent the moral degradation of American society by stigmatizing distilled spirits? - here's an interesting title from WB

> The United States of America has experienced a moral decline in recent years, straying from its Puritan values as a Christian nation and has become a modern day Sodom and Gomorrah. The nonsense of Gen Zers on Tik Tok, to the terrible finale of GOT, to the election of President Kardashian to the oval office, all share a common link: alcohol.

@lyxal Mmm, sounds like a threat

Use our Intertubes if you don’t want an...accident to happen to your precious blue-screened laptop

@hyper-neutrino WB né’ er disappoints

Oops

@NewPosts Indonesian gambling spam

hmm I got beaten to it by quite a bit, wasn't looking at my browser

i don't think it's spammy enough that i'll force spam nuke it tho. seems like someone asking an actual question that happens to belong nowhere on SE but since they don't advertise a product i don't think it fits SE's (IMO too limited) definition of "spam"

SE chat rate-limiting is really annoying

when i try to send like three messages within like 6 seconds

it makes me wait like 5 seconds half the time (and this edit made me wait 7 seconds)

What is chat rate limiting

It is what you get if you hit 1<enter>2<enter>3<enter> quickly

@wasif it's actually annoying, and means that the message doesn't send until a cool down period is finished

I see

@Bubbler nice explanation

chat rate limiting obviously makes sense but it's kind of annoying when i try to send a bunch of copy-pasted messages that i wrote up to send together without getting interrupted (multiline messages hide text which i sometimes don't want)

kinda wish mods got nerfed/no rate limit; i only really send long chains of messages because of moderation tasks. but i guess that might be seen as unfair xd

CMC: Make a 8BIT Computer on a breadboard and send me a picture.

@PyGamer0 that's too broad for a CMC

what about a 2 bit computer in Minecraft

define "computer"

the room is quiet!

No need to talk when you have nothing to talk about

imagine talking and not just lurking here like me

'Course, it's just been crazy the last dozen times I've checked in

@PyGamer0 i made one but my school's education edition subscription expired

When did the renewed activity start?

There was a long while where it was pretty dead

And Vyxal is alive!

#democracy

More or less last summer. Does anyone have a sense of when it became so popular again?

November, then maybe april

Was active in November, and then again in April?

As I remember, this summer the room was just in a state of anti-freeze

is it not just the start of summer right now

Ouch. How time flies

I meant last summer...

I'm now realizing what a waste of a year this has been...

insert 2020 on drugs quote here

and the drugs period is still going I guess

@AviFS sometimes silence is good

Be at peace...

RIP TNB

CMQ: Is log base a of b useful?

in code golfing

ye

natural log?

i don't actually think i've used log much in golfing

it's one of those things you could get away with using a digraph for

overload it onto a dyad

how

already done

github.com/Lyxal/Vyxal/blob/master/docs/elements.txt#L14

implemented it as mL and ml

oh

i thought you meant overload its arity somehow lol

no silly, I meant by types

:c

@lyxal how often do you find youself using ß in Vyxal

@PyGamer0 Enough to justify its existance

@PyGamer0 of course

add ln as a seperate builtin too

By the by, how's Dennis? Has anyone heard from him?

@AviFS No, so nobody knows.

You mean because he hasn't been active in TIO either?

I think he has been inactive online.

Hello!

Hi!

Recently the Mathematica certificate/license broke and someone mentioned it and it got fixed pretty quickly - so I believe Dennis is still well enough to keep TIO functioning as it always has but we haven't heard anything from him directly unfortunately

Glad to hear that, but sorry to hear the other. I hope his health is recovering

If a+b×i is a "Gaussian integer" when a and b are integers, why isn't a+b×i a "Gaussian number" when a and b are non-integer reals?

@wasif done

> The ring of Gaussian integers was introduced by Carl Friedrich Gauss

I think it means complex numbers were known before that, or Gauss was not interested in non-integral complex numbers

(quote from Wikipedia on Gaussian integer)

I wrote a note in TIO.chat as he was recently on, if others would like to sign!

@Bubbler So there are no other names for imaginary and complex numbers than "imaginary numbers" and "complex numbers"?

Just with something like '~ Name', if you'd like to, small gestures can sometimes make a world of a difference!

@AviFS I understand your good intentions but to be completely honest I would probably advise against that - Dennis's mention of his health was over a year ago, and the community has already wished him well (to which he's responded) and intermittently people hope Dennis is doing well.

I can't speak for Dennis, but I'd personally rather not have a health condition repeatedly brought up (it's a sensitive topic so I don't know how any other person would take that, of course), and I'd certainly not rather someone draw attention to it that much.

Also, I typically (and would advise others to) reserve pinging Dennis only for major or significant issues that require his attention as soon as he is able to provide it, like one time TIO entirely became unusable, which he fixed quite quickly.

Agreed, and I think Dennis would prefer if you didn't message him the second you see he's online. Let him return if and when he wants to, at his own pace.

@Adám I think so. Wikipedia says that the name "imaginary number" was introduced by Descartes, so I guess there's no reason to have another name for that. It doesn't say much about "complex numbers" though

@AviFS how to sign there

just write ~ then my username?

That was my idea, yeah

ok

(Still writing my response to your guys' statements from before.)

Hi!

It's been a while since I checked in here

How're things?

\o

not bad

○/

could be worse

@RedwolfPrograms o/

Did I miss BMG or is it tomorrow (I'll miss it either way)

Pretty typical, I don't recall anything exploding here recently.

BMG is in 17 hours

@RedwolfPrograms no you didn't miss

I see you changed your pfp

@RedwolfPrograms yeah

Are the bots running on my instance or HN's?

both, but yours aren't posting

@hyper-neutrino I see your point, that makes a lot of sense. I suppose we'd all want differently, and such interpersonal things are often vary greatly from one person to another. Personally, occasional notes of appreciation, and being reminded that I'm enthusiastically welcomed really help. Especially as I often find it hard to hold on to, and start to build my own mental barriers about everyone being mad and it being shameful if I showed up, which it sounded like he might also struggle with.

It’s definitely tricky though, and I totally see your point. Such things can definitely backfire and only heighten a sense of guilt, shame & pressure. I guess I really was just doing what I’d like done for me. I’ll certainly keep your guys’ perspectives in mind and refrain from doing so again any time soon, barring any new thoughts on the matter.

Both instances repond to status pings, oddly

I'm so stupid

I used setTimeout instead of setInterval, so the keepalive stuff for the socket that watches for new posts would only run once

Which meanie starred that! :<

@RedwolfPrograms congratulations

Okay, should be fine now.

@RedwolfPrograms no, you're Redwolf

alright, my instance is off now

Sorry for all the weird issues with the bots, my current implementation is a complete mess, once I get back I'm going to reimplement it with my library and harden it a lot better so it's more reliable

well, setTimeout would have worked if you include a setTimeout call in the handler itself, but :/

It'd still be less reliable though because an error in the function would prevent future iterations

You can technically put setTimeout at the first line

That's true

I wonder how that's handled for async functions

Just did a tiny bit of testing, that would behave more similarly to setInterval with async functions, whereas putting the setTimeout at the end would guarantee the function finishes before the next one runs

Putting at the start also guarantees that the current function finishes before the next one

There's no context switching in JS, so each scheduled job always runs to completion

For async functions that's not really true

For normal ones there's no difference yeah

@RedwolfPrograms For async functions, the task is swapped out on await until it is ready to continue, but each "section" between awaits is still guaranteed to run in one go

Of course, yeah

@RedwolfPrograms how is it living in yesterday?

Just four days until I'm home

So I take it y'aint enjoying it

The food here is either ridiculously overpriced, an hour away, or you can't get a reservation for a week. Almost every meal has been granola bars.

Oof

When covid is over, you should try coming over hear for a holiday

Be cool if you did