The Nineteenth Byte

Aiden Chow

00:00

dude actually im staring at my code and like i see what is wrong (it go 0+1, (0+1)+2, (0+1+2)+3, (0+1+2+3)+5, (0+1+2+3+5)+8, and so on), so maybe i dont start with 0 and 1. but then idk what i can starting with ahhhhh

Dannyu NDos

0

A: Generate /* line number comments */

Haskell, 55 bytes unlines.zipWith(\n l->"{-"++show n++"-}"++l)[0..].lines

This is the third time I've ever golfed.

Aiden Chow

@DannyuNDos ermmmmmm this a bit awkward, but theres literally a deleted answer with that exact code and some comments saying its wrong in that deleted answer

Dannyu NDos

Why tho?

Aiden Chow

im gonna copy paste the comment here and idk maybe u can understand it:

This doesn't work - different length line numbers break indentation. Test case: "main = do\n print 1\n print 2\n print 3\n print 4\n print 5\n print 6\n print 7\n print 8\n print 9\n print 10\n"

Dannyu NDos

Err

Lemme fix this

Unrelated String

00:12

don't bother, that spec is a disaster lmao

Wheat Wizard

@JPeroutek Radiation hardening is to vague to say for sure. Byte removal is a common but not the only "radiation". I would guess "Yes" but it depends on the wording of the challenge.

Dannyu NDos

Done.

Aiden Chow

01:14

oh wow now my programs r outputting 1 2 4 7 12 20 (differences 1 2 3 5 8), thats just fantastic

hmmm but add one to each gets 2 3 5 8 13 21

not quite what the challenge ask for

:((

Aiden Chow

01:39

ooh wait a fucking minute i think maybe i need to start at -1 1, cuz then 1 0, 0 1, 1 1, 1 2, 2 3, 3 5, and so on. result is 1 0 1 1 2 3 5... cumulative sum 1 1 2 3 5 8 13... exactly whats needed

oh wow ig i unknowingly stumble upon this sequence: en.wikipedia.org/wiki/…

Aiden Chow

02:42

0

A: Repeat your program to print Fibonacci numbers

Piet + ascii-piet, 53 bytes (15×7=105 codels), no header/footer ttlddtN ub?fnNssSkKCVrTtTLcfkdtlElt?l??T lL sdD S?? Output as unary. Try Piet online! Try Piet online! x2 Try Piet online! x3 Try Piet online! x5 Try Piet online! x10 This is so bad but my brain is completely fried rn, this will ...

ayyy lets go

my brain is actually mush rn

i thought it wouldnt be too hard at first but the number keep coming out wrong

bubbler probably gonna come out with some 30 byte answer after seeing this now lmao

i wonder if i should even try to put an explanation there so maybe other ppl can see some golfs

hmm maybe i should do a vertical loop instead of a horizontal one to output the 1's, was trying to do smth like this thinking it would save bytes but now looking at it im not too sure about that

i think mainly becuz the DP+ isnt at A1->A2 but instead further into the loop, which make some black codels

eh too lazy to think about it now

wait maybe i should print before doing the adding/swapping logic

would that save any byte...

Bbrk24

03:08

> if an inactive #if swift block contains unmatched parentheses, the #endif may be followed by an ill-formed string literal

never thought this bug would be useful

Bbrk24

03:22

@AidenChow Piet and Repiet were on a boat. Piet fell off. Who’s left

(sorry that popped into my mind and I had to say it)

Aiden Chow

@Bbrk24 i have no clue

Bbrk24

03:45

I’m referencing a joke that people used to annoy each other with when I was a kid (though of course it was spelled Pete and Repeat). The idea was that the other person would answer “Repeat” and then you’d just repeat yourself

user

Huh, I've been pronouncing it Piette

lyxαl

Pie it

Two syllables

Or pee et

Or pee it

All 3 are interchangeable

That's how I read it anyway

Aiden Chow

04:01

@lyxαl ive been pronouncing it this way

well its a real person name

bruh its pronounced like pete

ATaco

"peet" (Like Seat), as that's how you pronounce Piet Mondrian's name

Aiden Chow

damn learn smth new every day

ATaco

Basically, Piet-er Griffon is a thing.

Aiden Chow

errr..... just no

now im picturing peter griffin making abstract art lol

Bubbler

@AidenChow ~~you underestimated me~~

Aiden Chow

04:12

@Bubbler well fuck

shows how bad i am at piet lmao

Bubbler

hehe

Aiden Chow

i have severely overcomplicate it

Bubbler

"if something won't work, it is no-op" is doing great here

Aiden Chow

bro 1 0 roll so smart

if u look at my code i was tryna take inspiration from other ppls codes that have comments, so the first two line to the right i have is ran once but ignored for the rest of the time, 1 0 roll is so much smarter tho

Bubbler

ah yeah, that surely works but messes up the flow badly in piet

Aiden Chow

04:17

and also i was thinking in unary so i have this huge loop to repeat print

@Bubbler yea well i manage to make it work, and i was like whatever if it work then it work :\

also u dont even wanna now how long it take me to figure out the numbers correct to output proper amount

one second it output 1 3 6 11 19, next moment it go 1 2 4 7 12 20

and im just sitting there like wtf is going on

lyxαl

05:26

@user how come you always manage to shoehorn your way into my uni courses?

user

I am an integral part of the internet

lyxαl

This is like the 4th time you've created a password

in the one powerpoint

this is getting out of hand

mousetail

05:43

Wait till you head about pepper

Adám

06:08

Spice!

user

Chef to control it all

Neil

06:53

@lyxαl step 2 is wrong, surely?

I mean, even NTLM doesn't send the password across the network

lyxαl

07:17

@Neil tbf it's a summary diagram of a general process. It's not meant to be in depth correct, just enough to demonstrate a point

W D

07:46

Looking for sandbox review

3

A: Sandbox for Proposed Challenges

Could it be a possible chess move? code-golf decision-problem chess Backstory Professor Chesster has a library of chess puzzles with solutions, but a meteor has struck (not real, I promise) and now their hard drive may be corrupted! Luckily, Professor Chesster's puzzle FENs are stored securely, b...

# Summary
Check if a move inputted with algebraic notation could be valid. Return a truthy value if it is, and a falsy value otherwise

bruh there is no chat markdown lol

mousetail

@Neil Password is still normally sent over the network, only protected by SSL. Client side hashing is a thing but not widely adapted

It would require sending the salt to the client every time

emanresu A

And you have to hash it again on the server anyway, because you can't trust the client side

mousetail

In this case the server wouldn't even know the password at all so couldn't hash it

In client-side hashing, password is never sent. Server doesn't know it. If the client doesn't hash according to the specifications they are only hurting their own security

Client can choose to use a different hashing algorithm than the server expects but the server has no way of knowing

emanresu A

08:02

The point of storing hashed passwords is that even if the database is compromised people still can't gain access to accounts

Whereas if you know the hash of a password you can just send that, bypassing the client-side hashing

mousetail

Server could perform further hashing on the hash

Client could perform 100 iterations then server another 100

emanresu A

Although... if the risk of sending a password over a network is that it could get intercepted, surely it's just the same with a hashed password?

mousetail

If you use the password in multiple sites it's safter

emanresu A

True

The Empty String Photographer

C seems to have a hobby of segfaulting: tio.run/##S9ZNT07@/z83MTNPQ7O6oLSkWMNQ07r2/38A

mousetail

08:11

C stands for "Cegfault Central"

The Empty String Photographer

Also, ChatGPT found it difficult to understand the code.

mousetail

08:45

I need to delete a single word from a PDF, why is every program to modify PDFs ridiculously expensive

Neil

09:05

@mousetail which directory service are you using?

mousetail

I don't know about directory services specifically, just how authentication over internet usually works. I guess you can use a private key but that's not what the diagram suggests

W D

09:41

3

A: Sandbox for Proposed Challenges

Could it be a possible chess move? code-golf decision-problem chess Backstory Professor Chesster has a library of chess puzzles with solutions, but a meteor has struck (not real, I promise) and now their hard drive may be corrupted! Luckily, Professor Chesster's puzzle FENs are stored securely, b...

Bump

Summary: Check if a move inputted with algebraic notation could be valid. Return a truthy value if it is, and a falsy value otherwise

W D

11:32

Bump

Neil

@mousetail do you mean that or do you mean over https? I'll agree almost everyone uses basic auth over https

mousetail

Yea plain password over SSL (not necessarily HTTP)

Sandbox Posts

0

A: Sandbox for Proposed Challenges

Print every Unicode character code-golf That's it. Couldn't find any duplicates (surprisingly) so I'm hoping to be proven wrong.

New Posts

17:12

0

Q: Decode Chess Move

As the title says, given a valid chess move in algebraic notation, produce the string describing the move. Example: fxg8=Q+ -> Pawn takes g8 promotes to Queen check Bb7# -> Bishop to b7 mate Here is a list of all the keywords and their respective notation R -> Rook x -> ta...

Sandbox Posts

0

A: Sandbox for Proposed Challenges

Make a braille cribbage board The braille characters lie in the range U+2800 (⠀) to U+28FF (⣿), making it a SBCS (Single Byte Character Set). Oddly enough, the braille patterns look a lot like holes in a cribbage board. Here is an image of one of the many cribbage boards for reference: So you mu...

Rydwolf Programs

18:05

@mousetail And PHP stands for PegHault Pentral

mousetail

18:44

I thought PHP stood for "Pretty Horrible Phesign Phoices"

New Posts

0

Q: How Many Staircases

Problem You're a staircase engineer on a house and realize you only have n rectangles to create a staircase. So you are tasked with finding the number of ways to use n rectangles to tile a stairstep shape for the architects. Your task is to write a program that takes a non-negative integer n as i...

code-golf sequence code-challenge recursion

mousetail

19:43

Interesting that there are so many different ways to get to the catalan numbers

ais523

20:19

@cairdcoinheringaahing This explains a lot, but I think it may be causing some issues – recently I spent a couple of days writing an answer over 30KB long, and it got upvoted less than two minutes after being posted. Presumably that's because an upvote is one way to clear a review task? But there seems to be a real risk of, e.g., upvoting incorrect answers if people upvote before fully reading them.

caird coinheringaahin g

That is an issue - people should be more happy to "Skip" reviews if they don't want to review them

Rydwolf Programs

@ais523 I think part of it is that people know your answers are likely to be good before reading the whole thing

ais523

out of interest, do all my answers trigger review tasks? or is it only the answers on questions that are more than X days old, or that are particularly short or long?

Rydwolf Programs

I believe I got a Late Answers task from it

ais523

@RydwolfPrograms there are occasions where I've misunderstood the spec, and the answer usually gets upvoted out of habit anyway before someone (often me) realises it's wrong

Rydwolf Programs

20:22

So any answers on questions older than 30 days likely do generate reviews

ais523

in my first answer for the FGITW post experiment, I accidentally posted a snippet rather than a function; nobody noticed until I fixed it somewhat later

Rydwolf Programs

FWIW people tend to be fairly accepting of invalid I/O even if they notice it

ais523

it is at least usually easily fixable (the question in question was source-layout, making it potentially nontrivial to fix, but luckily the fix happened to be easy)

Rydwolf Programs

Personally I decide to vote based on the cleverness of the important part of the code, even if someone gets the I/O so egregiously wrong I bother leaving a comment, that won't change how I vote

Ah if it was restricted source or source layout that might change things a little bit yeah

ais523

ooh, tags in chat look like actual tags nowadays

I am still working on that TC-subsets-of-Jelly answer: finding new techniques for proving things TC is one of my hobbies, and it is doing wonders for suggesting directions for improving those techniques

golfing languages make concessions to practicality / the real world in ways that tarpits don't, and that in turn means that you get some subsets with some really weird restrictions (such as the base conversion operators working only in base 10)

oh, this is leading to another problem: I am getting close to the answer length limit, does anyone have suggestions for presenting the answer in a way that will fit within 64KiB?

one idea I've had is to post the relevant proofs on Esolang (or Wikipedia, if they were done by someone else and thus are acceptable there) so that they don't have to be contained within the post

I have already taken to editing the answer in an external editor, rather than directly on Stack Exchange, because it struggles with rendering 50KB worth of Markdown on every keystroke :-)

emanresu A

20:33

You're probably going to beat my 11 at some point

ais523

the hard limit in Jelly appears to be 15, assuming I haven't missed any relevant builtins – I could only find 15 different ways to create an infinite loop

emanresu A

Damnit chat

Not even utc midnight this time

ais523

but some of them are very hard to use; in particular, it is hard to remember data across the "call a function with no arguments" builtins, and I will probably need to move © and ® from their current subset into a subset that uses one of those

the duplicate message was probably a consequence of my Internet being terrible

emanresu A

I've been wondering if it's possible to do bitwise cyclic tag (or a variant of that) with just arithmetic on binary numbers (ideally without using a logarithm builtin)

ais523

I already have a working implementation of 256-Echo Tag using ⁹¹Hƭ€&xŒrṙß, but haven't posted it yet beacause 256-Echo Tag has not actually been proved Turing-complete (although it almost certainly is) – there's a compiler from tag systems but it doesn't fit within length 256

so I'm going to need to do some proving in order to get that answer posted

here's a TIO link that shows the general idea behind the construction

@emanresuA if you have enough arithmetic operators it's possible because ELEMENTARY + an infinite loop is Turing-complete

see, for example, the primality testing formulas

the question is how many of them you actually need

emanresu A

20:42

You have already used up most arithmetic stuff with blindfolded arithmetic, so it'd have to be almost entirely bitwise

ais523

it's Blindfolded Arithmetic that used up most of it, and I may need to remove that to split the arithmetic operators up

I already have ~Ḥ planned as a way to make arbitrary integer constants once I've run out of digits

although I haven't figured out what to use that for yet

so that uses up one of the bitwise operators, and the Echo Tag implementation uses up &

XOR is still available and probably is usable for something

by the way, if you're looking for tag-like Turing-complete systems that work purely arithmetically, see Spiral Rise, although it doesn't map onto Jelly very well

Sandbox Posts

0

A: Sandbox for Proposed Challenges

Minimal number of jumps to reach a square code-golf Given a Knight current square in a chess board and an array of unavailable squares, calculate the minimum number of jumps required to reach the target square. Input Knight square, array of unavailable squares, target square. Input might look lik...

emanresu A

Interesting, although it's a shame its builtins don't work well with golfing languages. You could potentially use a base conversion builtin for that last one, but you've found better uses for those

ais523

I still have one left! although I'm planning to use it to create a list literal rather than for computation (although maybe it's usable for both?)

also I discovered that ṣ can split an integer into its digits, like D does – it's a little janky but having a second D is probably going to be very helpful

emanresu A

There would be some interesting things you could do if Jelly had infinite lists

Although I suspect, like with my Vyxal answer, constructing those lists would be the hard part

ais523

20:56

it has #, which actually acts a lot like an infinite list in practice

this might actually be a fun challenge in Husk, but I don't know it well enough to really do well

emanresu A

It'd be interesting if ¡ and # worked when passed inf

ais523

for # you probably actually want to pass 1

as that gives you halting for free

emanresu A

Ohh true

ais523

finding a way to do halting is often one of the hardest parts, e.g. the Echo Tag implementation exploits a bug in the run-length encoder (it crashes when given an empty list, so I run-length encode and decode the list as a way to exit the program when the list empties)

or, hmm

emanresu A

I wonder if there's some construction for Turing-Completeness that relies entirely on checking the properties of a number (then # could be used without data storage)

ais523

21:00

maybe the bug is in the decoder?

yep, it's in the decoder

@emanresuA I can do it if I have a base conversion and a tag-like lookup table

but that's quite a lot of builtins to ask for

emanresu A

Could basically implement a tag system with that. That's probably a last resort if you run out of loops I guess

ais523

I might do it anyway, though, I have both of ḃ and y left, the hard part is just creating the data structure that y looks into

right, it's most of a tag system already (although it doesn't need a queue, which is a big advantage over tag systems when you run low on list manipulation builtins)

the idea is to use the Post correspondence problem

convert a number into digits in some base (ideally bijectively), map those digits onto substrings in two different ways, flatten both ways, are they equal?

emanresu A

Hm, neat

By the way, tag systems are still Turing-Complete if they halt when the state repeats, right?

ais523

yes, they can be adjusted for a few different halting states

"when the state repeats" is actually a particularly easy one to implement – explicit halt state and empty-the-queue are harder

(from the point of view of writing the tag system definition, I mean)

emanresu A

@ais523 In some of my solutions I've just added some form of output and defined a "halt state" from there (although print builtins are a bit limiting)

ais523

21:05

also it doesn't matter whether your requirement to repeat is "same value seen twice, possibly with other values in between" or "same value seen twice in a row" (although in most golfing languages the former is easier to implement, and it's also a somewhat easier situation to brnig about)

@emanresuA that's totally fine – but I have a self-imposed restriction on this one

actually one of the harder parts is just writing a nontrivial example program that halts to test the construction with :-D

especially when the halt requirement is someting awkward, like setting the internal value to 0 in Addition Automaton

Neil

22:03

...meanwhile, all TC Retina programs must contain at least one backtick...

New Posts

22:20

0

Q: Stevin's Notation

Stevin's Notation is a way to represent decimals in a non-fractional way. The Flemish mathematician and engineer Simon Stevin is remembered for his study of decimal fractions. Although he was not the first to use decimal fractions (they are found in the work of the tenth-century Islamic mathemat...

code-golf

DLosc

22:33

@ATaco Pip, 3 bytes: \,4

Transcript for