« first day (3790 days earlier)      last day (219 days later) » 

12:14 AM
@Bubbler The binomial transform looks pretty good. I don't understand it, but I'm sure others will
Wait, that's just finding the binomial coefficient with extra stuff. I'm not sure how interesting the extra stuff will be
Are there interesting ways to do a binomial transform that don't involve the same binomial coefficient code from other challenges?
12:46 AM
It is possible without calculating the binomial coefficients at all
Oh cool, I take back my second message then :P
I'm rebelling! Posting a question without running it by anyone. You'll be the first to see it in 3,2,1...
Q: Mr. Binary Counterman

AviFSMr. Binary Counterman Mr. Binary Counterman, son of Mr. Boolean Masker & Mrs. Even Oddify, follows in his parents’ footsteps and has a peculiar way of keeping track of the digits. When given a list of booleans, he counts the 1s and 0s separately, numbering the 1s with the odds & the 0s with the e...

1:39 AM
@cairdcoinheringaahing screw you and your "being smart" thing :(
that's a very clever solution tho good job +1
Now that Avi's edited in an explanation, it looks like we had the same approach tbh :/
oh, lol
well, if you get the same / better algorithm as OP, consider that a compliment ig :D
Good morning
@cairdcoinheringaahing I know, haha! I just commented that
@smarnav Morning!
Welcome to the Nineteenth Byte!
1:44 AM
And I was about to do the same thing in Jelly. Though I don't think I would've gotten it that short even doing the same thing
@smarnav Firstly, welcome to TNB! Does CBL happen to have any interpreter available yet?
You might have the winning answer, I wonder if anyone can beat 10. Nice!
Oh sorry, welcome @smarnav!
waiting for Jonathan Allan to get 7 or smth
I am new to code golf, and I was wondering if a challenge I was thinking about has any validity
or Bubbler with another funny base conversion thing :P
1:45 AM
Husk's builtin odd/even infinite lists might win
oh, true
yeah I was actually thinking husk probably has a much more direct way of doing this
actually, the enumerate into 0/1 list thing was something I was thinking about for the tic-tac-toe challenge. did that happen to inspire this challenge, @AviFS? :P
just out of curiosity
@cairdcoinheringaahing But not sure how you'd use that
@hyper-neutrino Darn, I've been caught!
How did you know?!?!
because I was thinking of something very similar :P
Haha, that's too funny
@AviFS Jelly, 11 bytes CÄḤ×Ɗ+ÄḤ’×Ʋ. Direct translation of your APL answer that's the same as my first approach :P
1:46 AM
Yeah, I was thinking for tic tac toe that it'd be very easy
two board configurations are enough to cover all even/odd cases in each possible slot (with rotation / flipping)
You compress a 9-long boolean list
yep :p
And then depending on whether you're given an even/odd number at an even/odd position
You negate the list
And then just fill in the numbers
yeah, something like that :P
1:47 AM
But it turned out I got really stuck on doing that last part well
It seemed like such a simple transformation, and yet...
So that's the origin story. And I could've used it to cheat on the other, once you guys optimize this part. I've been caught!
@hyper-neutrino Unfortunately not, since I have a lot of other work, and also I am questioning coming up with it, since making interpreters seems a lot more difficult than I thought. I think I still will try, but CBL might not look anything how it looks now when I am done. one issue that I am on the fence with is the 2d main array. anyway, I really hope to get the interpreter done, maybe over the summer.
@smarnav Ah, okay, I see. I was just wondering - we actually define languages by their implementation here (not by their specs, because interpreter bugs are considered to be part of the language features and behavior), so I'm not sure about the validity of your CBL answers as a result of that.
that is very true and makes a lot of sense.
@cairdcoinheringaahing Oh wow! So it might not have been so bad
How did you shave off the last byte?
Nevermind, I see the 10byte solution is significantly different!
I'll just go learn Jelly, and come back later haha
Was it a good challenge?
time to advertise again

  Jelly Hypertraining

Practice your Jelly :) To be a student, request access. Rules ...
@AviFS I found it quite enjoyable. Getting beaten by 33% was less enjoyable :P but that challenge was good!
1:54 AM
Ä (cumsum) vectorises at depth 1 (aka doing it on [A, B] where A and B are lists, it does Ä on both A and B), so I pair it with its complement, do Ä and (unhalve) on each, decrement elements in the first one and multiply each by the original elements. Finally, I take the column-wise sum
I specifically joined, since I was wondering if "Write an interpreter for a stripped-down version of COBOL in the language of your choice" would be a good idea, for a code-golf style challenge, since for some reason I am really interested in COBOL. I am kind of ok if this is a bad idea, since as I myself realized, interpreting is hard. please let you know what you think.
@hyper-neutrino if you ever have any questions about jelly or want to learn how to use it, feel free to request access to here :D
@AviFS Yeah it was fun :)
@hyper-neutrino Awesome! So glad
@hyper-neutrino Thanks, it's the worst...
I'm already a member
Or was...?
Yeah, still have access
Just hardly have done anything
oh yeah, didn't notice :P nice!
1:56 AM
Thanks! And thanks for the invite! I keep meaning and wanting to pick it up. And keep not...
@smarnav That sounds like it could be a decent idea, though note that "interpret this" questions sometimes become boring and less interesting than just tedious - if done right, it can be very enjoyable and offer a lot of clever approaches though. It seems too complex for me to just assess here, so I'd recommend putting it in the Sandbox for more feedback
It's one of the many langs on my learning list. But IDK when I'll prioritize it more
@AviFS haha, it took me quite a while to commit to Jelly and learn it too. It can be difficult. but since you seem to already have a good hang of APL at least the tacit concept won't be foreign, and learning chaining can be counterintuitive but it's consistent and you can just refer to the tutorial's chaining rules
the main part about any golfing languages is learning how a weird roundabout solution can have a shorter solution due to built-ins, and how you can use that to your advantage
Yeah, it seems super fun!!
J has been on my todo list for ages longer, and I haven't even learned that. It's really sad, haha
1:59 AM
I could look into learning J :P it looks pretty cool
I always want to look into learning more languages, but I've forgotten how my own languages work, so I never really bother :P
I get super excited and download a million books, read about it a bunch, play around with it meaninglessly through the first few bits of the tutorials, and then I'm onto the next craze, haha. Can others relate?
I've spent so much time with JS learning another language is impossible
I get that with Python + Jelly
2:04 AM
Joy is another one I've been really meaning to
They're distinct enough that I can remember how to code in each, but as soon as I throw in a new one I get confused :P
I just can't stand it when a language decides to do something sensibly instead of letting me commit horrific atrocities in order to do perfectly ordinary and sensible things
Also I like how if (new Boolean(false)) console.log("true"); works :p
hahaha, can't blame you there
we're all masochists at heart, after all
well, we are code golfers, so yes... :P
that's just what i meant, haha
2:07 AM
When you golf the code, the code golfs you
Only in Soviet Russia
(Does that joke mean anything to anyone?)
Yep (I'm old :/)
(Yes, from lots of time reading YT comments :p)
yep, I've heard the soviet russia joke plenty of times :P i don't recall if i've actually seen the source though
Oh, phew!!
I spend way too much time in YT comments, but never noticed it there!
2:09 AM
"In soviet russia, potato farm you" is one I found under a minecraft potato farm video :p
"In Russia, game plays you" is a voiceline in Overwatch :P
what the heck did I just watch? [non-explicit mild nsfw]
Honestly, if that's Lyxal's reaction, I'm not sure I want to click that
@lyxal Quality content, that's what
2:17 AM
@Bubbler o.O
Ugh, Jelly's zero indexing will make that long
what was the trick for grade down again?
since jelly doesn't have it (:()
I want to try this in Risky :p
I just thought of a really smart way to do it, actually.
@hyper-neutrino Grade up of boolean negation should do it
ah, alright. negate grade-up should work in general right?
i think that's what user suggested
2:21 AM
Just so y'all know, I'm disappointed at your lack of controversy for my proposal :)
Not in general, just for boolean input
@Bubbler You have a 0 in your output
@cairdcoinheringaahing It is allowed (which is written in the middle of the challenge)
Did not see that :P
@cairdcoinheringaahing Your output looks totally wrong
That's the intended behavior. You're understanding grade up wrong :P
Is broken?
@Bubbler Explain the 7 at the 5th index
wait wait wait
this is right
if you index that into the original list
it's sorted
if you want the permutation order
do double grade-up
Rank ascending is grade up twice, not once
2:28 AM
Then it's 10 bytes in Jelly ಠ_ಠ
This may or may not be why ỤỤ$ is one byte in my fork :P
note to self: add grade down and also rank asc/desc
@hyper-neutrino NỤ usually works
2:35 AM
What did you delete?
I found a super awesome solution in Risky but I'll have to double the length because of a tiny detail
So Jelly ties with APL :P
2:56 AM
@Bubbler Not if you have anything to say with it :P
Oh god I'm so stupid
yes what did you do this time
I was using the base convert operator instead of the xth item in array one
It's } not { D:
It works now!
44 bytes
Sounds painful
Definitely qualifies as an esolang, not a golflang :P
It's an insanely elegant solution though
I need to write an explanation for this
3:06 AM
The list of built-ins looks way too specialized for a language with only 16 ops (even with various overloads)
They are, yeah. It's not very well designed
Lots of the dyads are super specific because I had too many spaces to fill lol
You might want to learn something from the list of K built-ins (manual)
This was meant to be more a proof-of-concept, so I didn't really put too much effort into the built-ins.
Its a good idea tbh. A lot of languages die in the dev stage because creating a builtin list can be so tiring
A: Sandbox for Proposed Challenges

smarnavCOBOL(lite) interpreter. Your task is to code an interpreter for COBOL. This interpreter should use the free format, and cut a few corners, as long as a few test scripts will run on it. Notice; here I say interpreter, as that is the general direction of my inclination, however if it may save byte...

3:12 AM
@cairdcoinheringaahing I can testify to this.
writing the new list for the experimental vyxal branch has gotten a bit tedious
@hyper-neutrino I have just posted my question into the sandbox.I might make this question a bit more clear, later, but this is all that I have, now
3:40 AM
@Bubbler To be fair a significant part of the program is identity operators as padding, the actual logic is quite compact
3:50 AM
You know, golflangs spend zero bytes for padding :P
* most :p
They absolutely do not :P
At least people can't say Risky looks like I slammed my head into a computer keyboard several times, because the code page is small enough I would need a very oddly shaped head to do so
@RedwolfPrograms Doesn't that mean it has 88 instructions?
87, one's padding at the end
3:57 AM
Yeah, that's really not golfy :P
Keep in mind the byte count would be halved if I removed a single operator
This language could be really short in some situations, but they're rare (especially for longer programs)
4:20 AM
It could do significantly better if you find a way to change the program structure using all-ones padding in the middle
There's different ways you could have it determine the arities, which I suspect could make it significantly golfier.
I didn't want to make this language too complicated, though
Ugh, non-sandboxed-questions are the worst.
@Ausername the second one was sandboxed
What's wrong with the third?
4:31 AM
you can click OP and go to their meta profile and check their answers
I'm having trouble understanding Bubbler's impressive answer here
I think the algorithm is the same for both, but I'm looking at the APL one
(But this is more active than APL Orchard rn)
What is this train (⊢+2×⍋⌊⍥⍋⍒) doing to the right of ?
4:57 AM
@Ausername Uh oh, I'm in trouble! The second two are both mine; the only ones I have, haha
What part is unclear? How can I improve them?
with +17/0, +11/0, +11/0 on those three, no close votes, no negative comments, and no serious clarifications (just minor fixes or request questions) i think all of them are fine
i don't quite understand A username's remark... ¯\_(ツ)_/¯
I meant to finish my part of the blog today :/
Thanks hyper. Yeah I'm curious to hear what it was. In a perfect world, I do think that the tic tac toe could be clearer. But I'm pretty happy with the binary one!
@RedwolfPrograms Oh? Which one?!
We're trying to start up a CGCC blog
I think they're both great challenges, btw :p
No way, seriously?!!
Haha, thanks guys! You're too good to me
5:03 AM
The first blog post will be about whether or not golfing languages are good for the site, it should be posted by wednesday
What's the website for the blog? We have to get this on the starboard
Esolangs in general? Or golfing langs in particular, however you'd define that?
The posts will just be meta questions, that way they can be featured more easily
@AviFS Just golfing languages. This site has a bit of a history of not liking golfing languages sometimes :p
Oh wait! So is each topic a question and the blog posts are different peoples answers to the questions?
Or one question for the whole blog, and all the posts go as answers to it?
No, it's just going to be posts as questions, although people answering them with their own ideas could be really cool
Or is each blogpost an entirely unique post.
Oooh, I see. That's neat!
I feel like we'll have to establish what acceptable answers are, and what structure they should take
I feel like everyone will have lots of thoughts and praise they'll want to share, and it could slowly devolve if the answers are just there, free & unused
5:07 AM
It wasn't part of the initial plan to have answers on it at all, but I suppose it wouldn't be disallowed
This was where it was discussed how the posts should be formatted
I mean, how bad is it to let answers be like blog comments? I guess it's not the end of the world.
I think comment type comments would probably be better as comments
But answers which go into more depth on the topic would probably be fine
keep in mind that comments are intended to be ephemeral/temporary
if some relevant discussion is of some notability/significance, it is probably better as an answer
By the way, if you want to contribute to the blog, we've got this post where you can suggest topics or sign up to help write them!
@RedwolfPrograms I see, thanks!
I do wonder though if it isn't important to establish answer format sooner rather than later as well
5:10 AM
I doubt it'll be necessary
Answers probably won't be overly common
The question itself will be the blog post
i think we can just apply common sense moderation here
You guys know better; I trust your judgement
And the worst that happens is it changes!
all the resignation posts didn't have standards for answers either and some had a couple of answers and some have been deleted (forget if it was self or by mod), so we don't necessarily want/need a strict policy. it might not be a bad idea though
@AviFS It's basically (⍋⍋)⌊(⍋⍒) with the two s refactored into an .
@hyper-neutrino Oh no, you're totally right! I guess the resignation post answers were totally appropriate
@hyper-neutrino For sure, just offering the idea!
5:15 AM
hello everyone
That's not how you spell hi :|
@RedwolfPrograms H!
@Bubbler I thought I was confident with trains, but I'm not feeling that way right now!
does anyone know a golfy way to pop multiple elements from a list at once in python
Do you mind writing it as a defn?
5:16 AM
@Wasif do you have context to this
@hyper-neutrino yeah I am trying to solve the recent tic tac toe challenge and came up with an (non-brute force) algorithm
there might be ways to shorten it depending on what you are trying to pop, the nature of the list, etc
hey @Randal'Thor!
@cairdcoinheringaahing Was user 30? Either you know something I don't (in which case I'm interested) or you're getting him mixed up with someone else.
@hyper-neutrino Hi!
I saw SFF election chatter and joined the room :-P
yeah, i mentioned it here and we had a short discussion over it :P
which got derailed by a conversation about voting methods
what is SFF?
5:20 AM
In which I came up with several quite bad voting methods :p
@Wasif Science Fiction & Fantasy
and then since catija herself joined, we instead moved on to a meta discussion, lol. so yeah, long story short TNB has trouble sticking to a particular subject/topic
yes we went from breads to IDEs someday before
@hyper-neutrino Moderators aren't totally powerless on the election page; they can delete comments like on normal posts (although they can't delete actual posts/nominations, only CMs can do that).
@hyper-neutrino That's the nature of chat, innit.
@Randal'Thor oh, huh. all i know is nominations can't be touched and nomination flags aren't shown to us mods i think. forgot about comments, lol - that makes sense.
how do you create image links?
Like @Randal'Thor made
5:22 AM
@Wasif just type the site URL
Oh awesome feature


Where you can play with chat features (except flagging) and ch...
SE did a great job ensuring the contrast was high enough for the text to be visible, of course :p
@hyper-neutrino Nomination flags are kinda-sorta shown but not directly on the mod dashboard.
@RedwolfPrograms It works better in the black background of SFF chat :-)
@Wasif Does slicing count? Eg. arr[3:] gives a list with the first three elements 'popped' & arr[:-3] gives the list with the last three elements 'popped'
5:24 AM
@AviFS nope, i want to strictly pop it
@AviFS Well, it's just {⍵+2×(⍋⍋⍵)⌊⍋⍒⍵}
Hmmm... arr = arr[:-3] that's the best I can think of
Pyth might have something
@Bubbler Thanks! That I can play around with. Off to parse this blackmagic
@Randal'Thor huh. well, perhaps i will eventually see a nomination happen during my time moderating :P though i doubt we'll get nominees needing flagging
our elections are... let's just say, significantly less eventful than yours on SFF :)
@hyper-neutrino You'd hope.
Well, the current SFF election is significantly less eventful than (a) I was expecting or (b) the last one.
true. nothing seems to have exploded yet
5:26 AM
Hopefully with a less controversial result than the last two.
What happened those times?
There's at least 4 very strong candidates that I'd be happy to see as mods.
Ordering my top picks has proved to be very difficult; I haven't even voted yet :-D
@RedwolfPrograms In 2015, a mod was elected who ended up being asked to step down a few months later after a lot of controversial chat incidents. In 2016, an existing mod and an ex-mod deleted their accounts over distrust of a newly elected mod.
That definitely does sound a lot more eventful
Ooh, we now have diamonds from four different sites here
@RedwolfPrograms I've not renamed it, so it's still Attempt This Online, but RTO is indeed my plan
@RedwolfPrograms Three plus a CM?
5:32 AM
There's still a diamond so I still count it :p
the anime.SE main chat room also seems to have a pretty substantial mod presence. it's usually just three sites but there are fewer users so it's often like 50-75% mod
The old SFF chat room used to have so many blues that the local bot had a special "mods" command to make it report on how many mods and CMs were in the room.
interesting lol
Omg @Shaggy is this real
Given a long list of English words, how can I compute a subset of the alphabet that is sufficient for x percent of the words?
x is in the input along with the words
5:38 AM
pretty sure it's np-complete
I want the smallest subset
but if the alphabet is relatively small, you can just try everything
@JohnDvorak oh?
@JohnDvorak try all subsets?
@Anush reduces from 3sat
@Anush "does a suffice? b? ...z? ab? ac? ... yz? abc?"
@JohnDvorak sounds painful if you have lower and upper case
5:42 AM
You could try branch and bounding from the full alphabet.
@JohnDvorak nice idea
greedily chop off letters that leave the most words. That won't guarantee optimal answer, but it's really fast and it gives a good estimate
yeah, that's what i was thinking for a more efficient sol
X=100% is the set covering problem which actually came up earlier today
and it is NP-complete
@JohnDvorak I think I will try that
@hyper-neutrino interesting!
Maybe it should be a challenge
wait, nvm i think this is a different task
this isn't set cover
6:17 AM
@AviFS There's nothing specifically wrong with either, it's just a good idea to post in the Sandbox for Proposed Challenges to get feedback before posting.
@AviFS For next time, even when apparently quiet, people are actually sitting ready to answer stuff like that.
6:56 AM
In python, how do find the key in a dict with the maximum value?
@Anush lambda x:max(x,key=x.__getitem__)
er, i'm dumb
lambda x:max(x,key=x.get)
7:18 AM
Very nice
I have gotten another chance to post a python answer
7:36 AM
@Wasif you just need to ninja xnor, hyper, ovs, ophact, Jakque, dingledooper and possibly bubbler
or hope that they post in some other language
we're too busy outgolfing each other in APL/Jelly
thats a miracle they all didn't post python answer to the simple challenge
you need to count Noodle9 too
well, they are - I stopped trying :P
And I was asleep the whole time
I didn't post a python answer to the counterman one because i thought it'd be too easy TBH
7:37 AM
And in morning got the chance
that challenge is actually significantly easier for a procedural approach than a functional/golf lang approach
yeah it's very nice for an imperative lang(like Stax)
@Wasif ah yeah
I've kinda given up on golfing JS - It's far harder to golf than Vyxal, and far more likely to have an answer already :P. Also there's tsh and Arnauld - I've outgolfed Arnauld once, tsh never.
so many python golfers
I golf in whatever lang i found interesting
7:40 AM
@hyper-neutrino my last APL answer was in May, wow
@Razetime its unbelievable i ninja'd them nearly 60 times
power to you lol
@Razetime mine was in april lol
I have one python answer
After bounty ended APL people ran away
But is the bounty ended?
I think questions have run out
Razetime, Bubbler have still so many remaining bounties
7:41 AM
153 Vyxal
there are still many, many, many questions you can claim the bounty on
my last APL answer was never
25 splinter
98 deadfish~
@Ausername oh no
I have an optimal deadfish~ generator
Which takes `s` into account
haven't used it yet
I only claim 200s now
7:43 AM
36 JS and the rest are in whatever language I find interesting.
@Wasif I have a suboptimal deadfish~ generator
Which doesn't
@Wasif nice
@StackMeter though I didn't invent that
oh ok
Example use, outgolfs @Ausername by 44 bytes
The golf amount might grow bigger when used on big strings
are you sure this is optimal
7:54 AM
also, I have been reading about Python optimisation
The generator is written in C++
Uses priority queues
Where I have got this is a secret
well PQs are a pretty obvious thing to use. actually now that i think about this this probably isn't that infeasible to be perfectly optimal
seems rather difficult to optimize the {} and () usage though
hm yeah idk how i'd go about writing an efficient optimize for this
you can just brute force but ofc that's dumb
@hyper-neutrino you also wrote a generator didn't you
not for deadfish~
base deadfish is trivial to optimize
7:57 AM
oh sorry it was for deadfish
what do you get if you put ABCDEFGHIJKLMNOPQRSTUVWXYZ into your deadfish~ optimizer
@hyper-neutrino () isn't used at all it is important in 2D deadfish tho
let me check
right but you can technically terminate loops earlier with it in some strange circumstances

« first day (3790 days earlier)      last day (219 days later) »