« first day (4503 days earlier)      last day (639 days later) » 

00:26
@Hello,World! "Golf your solution by talking faster" is fantastic
7
@ATaco Haha, like that guy who used to record TV adverts and sang Bad for a talk show appearance-- I think he holds a world record for talking fast
Oh wait, more recently, he was on Great Big Story
 
1 hour later…
01:48
@ATaco You're welcome 🍜
@TheThonnu Hey save some for me >:(
02:36
1
A: "Hello, World!"

ceilingcatPA-RISC 1.1 machine language on HP-UX 10.20, 53 bytes 00: e8 20 00 00 bl 400010c8 <08>,r1 ; load address of label 08 to %r1 04: 34 1a 00 02 ldi 1,r26 ; branch delay slot, fd=1 to %r26 08: b4 39 00 3a addi 1d,r1,r25 ; pointer to string to %r25 0c: ...

 
3 hours later…
 
5 hours later…
10:38
CMC A060296(n), except that you may output any value not in the range [1..6] for n=2
 
1 hour later…
11:51
@Ginger which user? Me or RARE Kpop manifesto?
@ATaco the speech recognition doesn’t work that well, so I highly doubt that that person will have an easy time.
 
2 hours later…
lgtm
@PlaceReporter99 I, for one, don't understand the challenge.
@Adám basically, you need to calculate f(x) for a given list.
So x is a "list" (i.e. ordered set), and [a,b,c] is a list consisting of a and b and c?
@Adám yes.
14:33
You've not defined that anywhere.
@Adám indexing works how you would expect it to in python and $x*$ is the length of the list.
@Adám should be fixed
I wouldn't expect snakes to do indexing, but "lists" don't exist in traditional mathematical notation (which you use to describe the task), so you need to spell that out.
what are lists? we just don't know
@PlaceReporter99 why do you use x[2] instead of x₂ for indexing?
@Adám my python influence
14:39
Snaky.
I'd use |x| for the length, rather than the custom x*
@Adám i've defined a list (as an ordered set).
@Adám I don't think that matters that much.
@PlaceReporter99 You should probably change [n,n] into {n,n}
Much better.
You can remove both parens in the definition. They are just noise, without adding any clarity, imo.
@PlaceReporter99 I think such mathematical formulas would tend to use 1-based indexing, and it'd make the definition simpler too (no need for "−1").
14:53
Maybe state the domain of the input's elements?
Technically speaking, the first case is redundant to the last.
4 messages moved to ­Trash
@PlaceReporter99 Suggested test cases: [0,0,0,0] and [3.14,-2.718]
@Adám It is not. The top of the sigma will have a lower value than the bottom, which I suspect is undefined. However, I am certain that $\log_0 n$ is undefined.
15:08
@PlaceReporter99 OK, fair.
@Adám second test case is equal to 3.1646617321198729888284171456748891928500222706838871308414469486...
Yes, that's what my implementation gets as well. However, it serves to show that input can be negative and non-integer (unless you want to restrict that explicitly).
@Adám It's defined, why not add it?
Some languages have only non-negative integers as native numbers.
OK, I say it is ready for launch now.
at first the challenge confused me but it was just me reading the mathjax wrong, it seems fairly clear
15:17
@Adám made an edit that only allows number types supported by the language.
Uh oh, then a language might get away with 0 and 1!
@Adám Breaks the loophole "Abusing native number types to trivialise a problem."
ok
0
Q: Logarithmic Incrementation

HippopotomonstrosesquipedalianIn this challenge you will write a function that takes a list (ordered set) containing real numbers (except for the empty list) and calculates $$f(x)=\begin{cases}1 & \text{if } |x|=0 \\ x_n+1 & \text{if } |x|=1 \\ \log_{|x|}\sum_{n=1}^{|x|}{|x|}^{x_n} & \text{otherwise} \end{cases}$$ where \$|x|...

15:22
@PlaceReporter99 Is the [1,3,3,7] case really correct? I get 7.0057… rather than 7.0223…
@Adám let me recalculate
@Adám correct
15:53
0
Q: Output endless powers of 2

DadsdyIn any programming language, make a program that outputs infinite subsequent powers of two, starting at any power of two with finite integer exponent. Note: it only has to theoretically output infinite subequent powers of two given infinite integer size (so you can ignore any potential issues wit...

 
2 hours later…
18:16
0
Q: Can I host a competition in multiple locations?

DJMcMayhemI'm interested in running a KOTH in the near future. I would like the competition to be open to as many people as possible, however I know some people that uninterested in posting answers to Stack Exchange. But I'd also be happy to run it here as well. So I'm wondering if I can host this challeng...

0
A: Sandbox for Proposed Challenges

97.100.97.109Longest Tiles Combo The New York Times has a puzzle game called Tiles on their website. This game consists of a grid of tiles, each made of multiple visual elements which are shared between tiles. The player first selects two tiles which share any visual element, removing those tiles from play....

18:48
0
Q: Friends at the Theater

Venkatamurthy NarayanasubramanA man wants his friends to go to a play with him. The man has N friends. Each friend has different popularity from the others. Friend i has a popularity score Pi, and the man wants to maximize the sum of all popularity scores of the friends joining him to go the play. Friend i will only join th...

i changed my display name for the site, how do i make it update on chat as well?
19:06
wait for it
0
Q: Ancient High-Fives

Johnathan NambaswamyIn ancient times, high fiving was a bonding experience for cavemen. However, the cavemen were selective about who they high-fived and didn’t necessarily high five everyone but only usually only the others they were close to. Giving a random number of cavemen (x) greater than 1 and a random numbe...

19:25
@Jacob change your chat parent site
CMC: Uppercase a word, but only the trailing lowercase letters e.g. McDonald -> McDONALD, DeVere -> DeVERE, normal -> NORMAL
@Neil Shasta (no interpreter yet), 29 bytes: {(replace/[a-z]*$/uppercase)}
but it's not valid since i haven't finished the compiler
\p{Ll}*$
\u0
@Neil JavaScript (Node.js), 42 bytes: s=>s.replace(/[a-z]*$/,a=>a.toUpperCase())
19:41
oh yes, QuadR can ofc use [a-z] instead of \p{Ll}
so 11 bytes
Yes, and here's 11 bytes in APL: ⌈@(∧\⍢⌽⊢=⌊) Try it online!
0
A: Sandbox for Proposed Challenges

DadsdyText to Bezier curve Bezier curves are a type of point-interpolation that doesn't lead to lines going up super high (unline cubic spline). Formulas are in the link. Cool renderer is here. Your task: Given a string as input, output points that, when rendered into a single bezier curve, is the stri...

hello
19:51
ok my username works now. from now on you will call me noodle man 🍜
or jacob i dont care
but professionally i am to be known as the former
20:14
noodle man 🍜
20:51
shame it's not using a publicly available model for the speech recognition
or else you could do something really funky to mathematically optimize the shortest possible input, and it would barely sound recognizable
 
2 hours later…
22:30
@UnrelatedString I feel like with a sufficiently optimized input you could get like, a millisecond of static to be read as the right program if it doesn't check the actual timing anywhere
22:44
i assume the practical speech recognition technology being used does check the actual timing :P
but conversely if it was being scored on bytes instead of seconds you could probably fine tune some extremely low bitrate audio to just barely produce the right program
@noodleman the noodleification is now complete
noobdles
You joke now, but once you try porting answers from one language to another, you too will become noodles
i lowkey want to actually make an esolang built on open-model speech recognition now :P
maybe even some weird majority decision making thing polling multiple models so there's even more jank to exploit
enforce some particular uncompressed audio format for the input, presumably one that the model(s) use natively, so it can be scored in bytes without having to worry about compression as another dimension of golfing
and make the syntax super plain-english-y so it feels like actually giving spoken instructions that then feel all the more horrifying when you listen to them optimized :P
23:04
even better: make the instructions IPA sounds
if hello world doesn't summon Cthulu you're doing it wrong
ahahhahahhahahahhaha
hel'lo ftaghn
make the interpreter some kind of praat plugin
private repo
ah oops
@UnrelatedString fixed
basically the idea was to FFT the sound and use the different freqs as different stacks, and different amplitudes are different instructions
sadly cant do it rn, i gotta finish up my phase 2 submission for the apl contest
Dang that's still an ongoing competition?
another month left
> The deadline for submissions is Friday 28 July 2023 at 23:59 UTC.
Wow
That's longer than I expected
23:34
ive got gold for every one in phase 1 except last
yeah, it's a super long-term competition
@lyxal yeah before you asked i thought it was june 5
Y'all getting several months while I had to speedrun my solutions in 13 days :p
last year?
This year
I do a little problem beta testing
23:35
hrm so you didnt have to speedrun them
tbh i did phase one in like 2 days
@Seggan Well technically I did because feedback deadline :p
one idea for golfiness measure of languages: use previous years' phase 1 tasks
im particularly proud of my phase 1 problem 2 solution
@Bubbler but they dont get shorter?
wdym?
nvm
23:40
@Seggan yooooooooooo
@UnrelatedString ima rewrite it in either kotlin or rust, depending on which has a (better) DFFT lib

« first day (4503 days earlier)      last day (639 days later) »