The Nineteenth Byte

lyxal

00:01

well duh

websites didn't exist 50 years ago :p

the first one is just over 30 years old :p

33 years to be precise

thejonymyster

lyxal

then write better jokes :p

50 is a number that seems like it's a placeholder big number

thejonymyster

34 years ago you couldnt find that on any web site

lyxal

see, much better

thejonymyster

Hell yeah

lyxal

00:13

for more tips on how to be funny, check out my blog post containing 10 tips on how to make the joke

Ginger

suspicious.

Unrelated String

This... this pains me to say, but you can golf ẆṚLÞ to UẆU. — Unrelated String 6 secs ago

12

emanresu A

uwu

Unrelated String

("substrings, reverse, stable sort by length" versus "reverse, substrings, reverse each")

lyxal

00:30

@emanresuA because it acts as a somewhat inverse of "is sorted". The naming is a bit bad, but it's basically "is sorted in descending order"

emanresu A

I'm aware, it's just ... oddly specific

lyxal

Aaron came up with the idea

github.com/Vyxal/Vyxal/issues/741

emanresu A

@UnrelatedString i did at one point have uwu in a version of this

lyxal

~~This is Joe Biden's code golf~~

Unrelated String

00:46

pfffff

amazing

noodle person

@UnrelatedString I have just realized something I might change in my sandbox question and I'd like your feedback

Unrelated String

alright

noodle person

Sequences of multiple ds are treated as chains of dips even if they don't end in i or p, so e.g. dddid is ⊙⊙⊙∘, dddpi is ⊙⊙⊙π

But, if there is only a single d, it's only treated as a dip chain if the next char is i or p

Something like dpi isn't allowed, but dpid is, and ddpi is allowed and parsed differently from ddpid

Unrelated String

oooh

That sounds like a good addition

noodle person

01:02

Ok. I'm not sure how to phrase it clearly

Unrelated String

I feel like you phrased it pretty well there :p

noodle person

Okay, I guess I'll just say what I said :P

~~(I'd rather be famous instead)~~

I don't think your suggested regex works anymore, and I don't know that it's possible to express clearly with a regex unless you use negative lookahead

(id|pi|dip|(?<=d)[pi]|dd+|d[pi])

Does that look right? I feel like it's wrong

Unrelated String

01:25

id|pi|dip|d[ip]|d{2,}[ip]?

?

noodle person

I don't think d[ip] is right

Wait no yeah

But d{2,}[ip]? isn't

because ddpi is dip dip pi

emanresu A

or just d+[ip]|d{2,}

noodle person

@emanresuA Does that cover the ddpi case right?

Unrelated String

@noodleperson It wouldn't match right greedily but I don't think that's automatically an issue?

noodle person

It wouldn't since the [ip] is captured, I am 90% sure it needs the lookbehind assertion

Unrelated String

01:29

@noodleperson it's just a cleaner version of mine :p

noodle person

That was a response to emanresu's

emanresu A

@noodleperson think so? there should only be one way for it to match in theory

... i'm confused

noodle person

ddpi would match d+[ip]

which is wrong

because then the i can't be consumed

emanresu A

isn't dipiddd ambiguous? can be dip id dd or di pi ddd

Unrelated String

uh oh

noodle person

01:31

It is dip id dd

emanresu A

then you need to specify what takes precedence

noodle person

but it behaves very very strangely

Well not that strangely I guess

dipidd is di pi dd

dipiddd is dip id dd

Unrelated String

...wow

emanresu A

so... what rules exactly govern that?

noodle person

Well, with dipidd, it can't be dip because idd is invalid

Unrelated String

01:33

I guess it wants to prioritize having the full dip mnemonic, so long as there's no unparseable single d

noodle person

Yeah

I'm going to ask Uiua discord people, though. I'm not sure

emanresu A

Also what about ddidd and ddpiddd

noodle person

ddidd is ddi dd

ddpiddd is dd pi ddd

Wait

I didn't check the latter I gotta see

It might be that ddpiddd is ddp id dd

Ok yeah its the latter

ddp id dd

emanresu A

so is it always greedily matching the longest prefix it can?

noodle person

I guess so

I will try to come up with some more robust examples now. Thank you for helping lol it's harder than I thought

emanresu A

01:47

@emanresuA (also it's still ambiguous even without the dddd rule - dipiddi for example)

Unrelated String

02:43

@emanresuA he used the other reverse 😔

emanresu A

noooooooo

rydwolf

03:03

So I'm on ResearchGate reading about hardware security right

Unrelated String

rydwolf

Ahh it's probably because the name of the specific impedance-based attack is "LeakyOhm"?

Unrelated String

PFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF

rydwolf

I am going to go ahead and guess this is the result of LLMs lol

Unrelated String

yeah

emanresu A

03:17

amazing

lyxal

04:00

@emanresuA is ÞSṘÞṡ not just the same as ÞSÞṡ?

Why reverse the list of sublists if you're just going to sort it?

Or does it somehow do some stable sort ordering thing?

Unrelated String

that

Earlier substrings need to appear later so the last one surviving the filter is the earliest one to be the longest

lyxal

I suppose that's what I get for not having a sublists that returns output in length sorted order to begin with :p

Oh the joys of infinite lists

Unrelated String

:p

in this case it's the same either way--this is exactly what Jelly has to do and Jelly does have it length sorted

lyxal

Yes but at least emanresu'd get a byte shave

rydwolf

04:19

I am currently learning just how powerful of a motivator a very long paper can be for doing anything that is not writing that paper

I have a 20 page literature review due in 57 hours

I have not yet decided what topic it will be over

And it will not be a subject I am familiar with

Unrelated String

ikr

(also good luck)

(how do you even find 20 pages of content to write about ;_;)

rydwolf

05:42

Stumbled across this paper from 2020, JS-based KASLR break is crazy

rydwolf

06:32

This is not a particularly useful way to utilize one of these diagrams

You can't just take the cartesian product of the axes lol

Well, they do at least make the bizarre choice to put faint, almost invisible acronyms diagonally in the background of two of the four quadrants, and the least useful visualization possible for gate-level masking

lyxal

07:55

I was working on solving that hidden double challenge and realised my solution didn't work for strings that aren't contiguous and I don't think I'll continue solving because my answer is a nasty mess

But to make the most of it, Try it Online!

38 glorious bytes of nastiness

⌈DDẏ¨V⟇Ṡ¨Zṗ$c*h;ÞGD£¤$↳VṄ:₌‡LI‡¥eƒr¥fȦ

thejonymyster

08:07

thanks for playing!

emanresu A

09:43

@emanresuA this turned out to indeed be a terrible idea

9

lyxal

12:15

ᱫ

that really looks like the "shhhh" among us character in the role reveal screen

somehow in the same unicode block there's also ᱷ

The Indians predicted among us in 1925

mousetail 'he-him'

12:35

CMegaC Recreate the entire among us game with unicode characters

I've been thinking about creating an among us KOTH for a while now, but even if I simplify it as far as possible, writing a simple bot for it would probably require more effort than KOTH submitters are typically willing to do

The idea is: Every turn, every bot chooses a room to go to, gets to see who else is in that room, and if they survive, can say one fact. Then they must vote who to eliminate

A fact is like "Green as in room C" or "Orange was not in room D", only very basic stuff allowed

The impostor, if in a room with at least one other player, could choose to kill them.

Ginger

13:44

@emanresuA this is extremely clever, well done

mousetail 'he-him'

14:21

Anyone have an idea for a simple decision-problem string that vyxal does not have a builtin for? For Byte Heist

Hello, World!

17:50

0

A: "Hello, World!"

How dare you fuck the brain, 166 bytes This is the problem with HDYFTB. Printing strings is really anoying II=+=+=+=+=+IIIIIIIIP=DDDP^IIIIPPIIIP^II=+=+=+=+^II=+=+=+DDDDv+P D)v D)II=+=+=+=+P=+^II=+=+=+v+IIIIIIIP=^=v D^Dv)II=+=+=+DDDDDDv+P=IIIP^DDDP^DD=+=+v+P D)II=+=+=+=+IPH Try it online

Gleb

19:39

hello i am back

thejonymyster

o/ salutations golfer

Gleb

so uhh

what now

guess i'll make the "set initial tape" argument for How dare you fuck the brain

Gleb

19:57

okay done
https://tio.run/##fVRNj5swEL3zKxyfcIEU1GoPIO9p06ZStFn10gPLSsR2ElQKFia7Req596r/sH@EjgETmiXl4mHmzXsz4w/Z1MeyeNdm32RZ1Ug1yuKS@lFaHSj8LGF9jv0w0Y6lLKXtE20qmVNYYj9ZgpnVNonqVAoKjjfBTcTF7nSgvvVyzHKBABhaKNujPoNS7HEc9pjABIBmiCkMaKTqKhsk4vdhLgq7/yFekHRassyKGjQQ6lU0RCeR2zEaaivmMqFlxbtgPMYScmZxaBBx6ehikO7eQuduLVZyQTHG7cMTWbdgREwPiIFv6C/Q9WZ7WyNjJnUTnzAZxXtekf@LuJsgvFnEEyBMVZex5y42n/cAMVlBWzY7VrbRIK4oONRMhgw0ybjHQ4IB/wdLsSnbCRJqEmaADpDSM9Qx0GjcFBlNmPwZCtKdhG62JmlBfdI7h30fO19Q/FhA62wylinZ2ytkbDLiKf7Pz19jBmLAboaEWZ6@wAkRKW8QF/usEByTaAeOrx2e0de9@rMKvycK9FLhVMyRj4ysP4Kvadc4NPDL0EdsZgd@w3TrwwymO6k/k…

mousetail 'he-him'

21:26

The "preview" Dolphin generated for a folder of videos that is still being downloaded

thejonymyster

pretty

mousetail 'he-him'

I offer a 50 rep bounty if you can identify the videos from the previous of the first ~1 kb

user

I posted this in the CGCC Discord too, but I'm proud of myself, so here it is again: github.com/ysthakur/fred

I made a programming language for exploring a compiler optimization for this very old cyclic reference counting algorithm called lazy mark scan

Annnnd the fuzz tests actually failed on the latest commit

Before that it was mostly Valgrind shitting itself in the CI for some reason

lyxal

21:57

Is the hamster lore real?

Based if true

user

22:13

@lyxal Sadly, no

lyxal

So sad

I cry eberytime

user

22:32

Just imagine that I truly did have a hamster called Freddie Krueger and that, like John Wick, I decided to seek revenge on the sick bastards that killed him

A tragic, powerful story

lyxal

@user "I once saw him kill 3 garbage collection cycles with a pencil. A fricking pencil"

Transcript for