The Nineteenth Byte

New Posts

00:09

0

It can be easily proven using Hall's marriage theorem that given fixed \$n\$ and \$k<n/2\$, there is an injective (one-to-one) function from all \$n\$-bit strings with \$k\$ ones to \$n\$-bit strings with \$k+1\$ ones such that an input and its corresponding output differ by exactly one bit. For ...

Adám

00:44

@Ginger o⍜o

ATaco

o◌̮o

Ginger

@Jacob ^^

Rydwolf Programs

Not really a curved line is it

Adám

01:01

No, sorry, I don't think that exists.

Jacob

@ATaco thanks for this

o◌̮o

>◌̮<

-◌̮-

;◌̮;

i stop now

Ginger

01:18

ಠ◌̮ಠ

AviFS

01:57

I've used this parsing generator several times in the past, but now I can't get it installed for the life of me. Does someone mind trying it on their end?

github.com/lark-parser/lark

It should just be one pip install lark away.

Does someone have a few minutes?

emanresu A

03:26

@RydwolfPrograms o͜○o

Rydwolf Programs

03:37

@AviFS Works for me

(Python 3.10.2 (main, Feb 25 2022, 22:40:35) [GCC 10.3.1 20211027] on linux)

From a Docker container, specifically

lyxal

11:39

@emanresuA this reminds me of the imaginary older brother from DHMIS Season 1 Episode 4 for some reason lol

Adám

14:09

Announcement: Schedule and registration for APL Seeds '23, the 22 March 2023 (14:00 UTC) free online event aimed at new APLers and the APL-curious (although all are welcome), is now available.

4

Sandbox Posts

14:47

0

A: Sandbox for Proposed Challenges

Elevator movement order There is a building with an infinite number of floors and you are the only person in an elevator that can lift an infinite number of people. You go home, but on one of the floors, several people enter the elevator and enter their destination floors. Your program should han...

UndoneStudios

16:28

Hi

Seggan

17:28

me is bak

DLosc

17:42

@Jacob I like this one best

The Thonnu

18:32

codegolf.stackexchange.com/questions/6487/… why is this getting close votes?

DLosc

@TheThonnu The two close votes at the moment are "Needs details or clarity." The thing I found most confusing about the question was whether the input array had to use the names of nuts given or whether it could contain some other values. It looks like the OP's intention was that any 11 distinct values could be used, but that was only clarified in the comments.

There might also be some confusion over the exact parameters of "Solutions may not simply shuffle the array until it becomes unique by chance."

Rydwolf Programs

Yeah, that's why I VtC'd

It's unobservable

(And I don't see why it needs to be a rule in the first place)

Sandbox Posts

19:54

0

A: Sandbox for Proposed Challenges

Help Me Type on My New Keyboard Overview As most of y'all who visit chat know, I am Russian by ethnicity[citation not needed]. Recently, I switched from a keyboard that maps the letters 1-to-1 (mnemonic) to Windows' default mnemonic keyboard. Here are the letter mappings: "`" -> "ъ" "q" -> "я" "w" -

user

> As most of y'all who visit chat know, I am Russian by ethnicity[citation not needed]

citation-needed

(for that whole sentence, not for you being Russian)

Seggan

im quite sure most of chat knows i am?

DLosc

@Seggan I do, but I mainly remembered it because I have Russian ancestry myself

Seggan

doesnt change the fact you knew :P

Rydwolf Programs

20:15

I knew you know Russian

user

@Seggan But you need a citation :P

I was just kidding, don't take that message seriously lol

The Thonnu

@DLosc @RydwolfPrograms it seems clear to me. OP just named the nuts so that we would know how many there are.

But it's closed now so no need to worry about it

Rydwolf Programs

I don't care about what the nuts are named

It's the "you can't sort it by shuffling the list" thing that's unobservable

@TheThonnu If anything there's much more of a need to worry about it now that it's closed

A question that's closed and shouldn't be is a much bigger issue than a question that should be closed and isn't

mathcat

@Seggan I knew you had a russian keyboard

Rydwolf Programs

So if you ever think a question's wrongly closed, it's always worth discussing

The Thonnu

20:20

@RydwolfPrograms oh that just means you can't randomly shuffle it until you get the answer

So anything that doesn't use random numbers is valid

Unrelated String

@Seggan i assumed you go to a russian orthodox church and by extension are probably ethnically russian but i don't really know how the various orthodox divisions work so there could be some other circumstances for a russian-speaking non-russian from, say, one of the other soviet remnant states

The Thonnu

Obviously I'm not OP though, this is just my take on the challenge

Rydwolf Programs

@TheThonnu Which is a) unobservable and b) a dumb requirement

You can objectively determine whether something's using that specific strategy, or something different enough to count as a different approach

mousetail

No randomness seams like a reasonable requirement

Rydwolf Programs

It doesn't to me, but even if OP for some reason wanted it to be, they need to say that not ban a very specific way to use randomness

mousetail

20:29

Basically there must be a finite worst case running time, this is measurable

Rydwolf Programs

Wait is this restricted complexity?

Jacob

maybe they should have labeled it restricted-complexity

Rydwolf Programs

No it's not restricted-complexity, at least according to the text of the challenge

And most answers use an all-permutations-of built-in which would definitely not be allowed if it was

I mean it's a moot point anyway since I doubt it'd ever be golfiest to use randomness, but if it was, that would probably be a cool enough strategy that banning it is unnecessarily restricting people's creativity

mousetail

IMHO shuffling should never be allowed, no known PRNG has enough enthropy to potentially generate every permutation, so they wouldn't technically solve the challenge

Rydwolf Programs

We pretend they do tho, right?

Just like how floats can't actually give the right answer most of the time

If randomness actually broke the challenge in some way, or trivialized it, it'd be fine to ban it. But there's just no point here; this challenge has nothing at all to do with randomness, and a stochastic solution would just be an odd, and thus probably creative and interesting, approach

I mean, allowing most language's built-in randoms is the same sort of hardware idealism

I'm pretty sure Math.random in JS doesn't exactly use quantum physics to get a truly random output :p

mousetail

20:34

This creates some odd paradoxes when implementing a custom RNG though. How good does a custom RNG need to be such that you are allowed to assume it's perfect?

Rydwolf Programs

A custom RNG can't be assumed to be perfect I don't think

Just ones provided to you

Seggan

@UnrelatedString nah im russian baptist

@RydwolfPrograms yep

@RydwolfPrograms and yep

mathcat

@RydwolfPrograms feature-request

mousetail

What if your language lacks a PRNG? Like codegolf.stackexchange.com/a/257346/91213 this answer does

Rydwolf Programs

I think the assumption tends to be that the hardware executes the algorithm perfectly, whether that's int/float precision, PRNG randomness, or call stack size (tho I think you can "opt in" to treating those more realistically)

But your algorithm gets no such special treatment

@mousetail I think that's technically invalid according to our rules

If a language doesn't have RNG it just can't compete, but enforcing that is less fun so we tend to not

(I think it's agaisnt the rules still to take random values from input, but that's another good option)

mousetail

20:38

IMHO if you implement XORshift 128 or Meraine Twister yourself you should be able to use it the exact same situations as the built in variants. In golfing this might not be relavent but in fastest-code it could be very wothwhile to implement it yourself

Rydwolf Programs

I think that'd be a case where you just ask OP to allow it on a case-by-case or question-by-question basis

Something that used to happen all the time

(which is progress IMO)

But also, you still need a seed for that custom PRNG

But either way, this is pretty tangential to the question in question

And it just goes to show how unobservable "randomness"/"shuffling" is

mousetail

I agree with closing that question btw, just saying my interpretation but the question should state it

Ginger

just found this in the RPi GTK theme file:

/*
   These wildcard seems unavoidable, need to investigate.
   Wildcards are bad and troublesome, use them with care,
   or better, just don't.
   Everytime a wildcard is used a kitten dies, painfully.
*/

Seggan

lol

its like that sudo make me a sandwich one

user

@Ginger Are they talking about wildcard imports or something else?

Ginger

20:50

wildcard CSS selectors

mousetail

Wait what's the issue with wildcard CSS selectors?

mathcat

@Ginger *starts using wildcards excessively*.

mousetail

I use *, *:before, *:after {box-sizing: border-box} in every CSS file

Rydwolf Programs

Seems like you could run into a lot of browser-specific weirdness

mousetail

That line is explicitly used to fix browser specific weirdness

So all other CSS rules have a neutral base

Often I'd add margin: 0; padding: 0 too

Wheat Wizard

22:04

@RydwolfPrograms The outcome being deterministic is observable and eliminates the shuffle algorithm.

Rydwolf Programs

Yeah, and it would be perfectly clear if that's what the requirement was

But you can have nondeterministic output without shuffling anything, and you can have deterministic output with certain definitions of "shuffling" or certain ways of doing it

Wheat Wizard

It says what the requirement is very explicitly.

"The sort employed must be a deterministic one"

Rydwolf Programs

Oh, right

I think the first sentence should be edited out but yeah, VTROing

Oh wait it's open

ATaco

@WheatWizard math.randomseed(0)...

Wheat Wizard

@ATaco Yeah that's one way to make it deterministic, but it costs a lot of bytes.

I'm not sure the shuffle method is even that good so I question why it is banned in the first place.

But this was 2012, and things were different then.

Rydwolf Programs

22:12

2 hours ago, by Rydwolf Programs

If randomness actually broke the challenge in some way, or trivialized it, it'd be fine to ban it. But there's just no point here; this challenge has nothing at all to do with randomness, and a stochastic solution would just be an odd, and thus probably creative and interesting, approach

Wait 2012? I thought this was posted earlier today lol

Must've misread which queue I was in

I could've sworn I saw it in the HNQs...

Wheat Wizard

In a way, I do think it's unfortunate that it is clear, because while the challenge is interesting at its core, it has some issues (of course it is from 2012, so this is all with 10 years of hindsight and acquired taste etc.). It would be nice if it could be closed and reposed with more modern rules. But I don't think it's a good idea to close a challenge that is clear just because it's not good.

user

Technically, we use PRNGs instead of real RNGs anyway

ATaco

Really though, shuffiling until truthy is identical to "Get all permutations and find the first truthy".

Wheat Wizard

Identical in terms of what?

ATaco

Depending on the language, the overhead of shuffling and obtaining all permutations is the same (See the Jelly answer in particular). Then, finding the first of the two which passes is roughly the same.

Rydwolf Programs

22:23

@user Not necessarily, I'm sure there's a praclang out there that has built-in TRNG support

ATaco

Would the response time of a ping be considered a source of true randomness due to the various imperfections of a line which may have a small but cumulative quantum effect?

Rydwolf Programs

I don't think it's an approved source of randomness but people would probably be fine with it

Wheat Wizard

Don't know for sure I'd say there's probably a 65% chance of it being true randomness and a 45% chance of it being pseudorandomness.

ATaco

**CMC:** Produce a TRNG number.

Wheat Wizard

8

14

Ginger

22:33

wheat wizard attracts stars the way gamers attract IRS agents

ATaco

Always relevant XKCD

user

Starbait is funnier when it's by someone who doesn't do it often

Ginger

nobody expects the mods to be funny ofc

q:

Jo King

ouch

5

user

It's okay, Ginger was just joking

Ginger

22:37

*rimshot*

dammit that's two jokes I should've come up with and failed to do so

Bubbler

22:50

@ATaco Serious answer: x86 machine code, 3 bytes, 0f c7 f8 = rdseed eax

ATaco

Don't forget a ret, otherwise it's a snippit :)

Bubbler

Well yeah, that's 4 bytes then

caird coinheringaahing

23:21

@Ginger ಠ_ಠ

I've been being funny in TNB since before you arrived, young one

Ginger

mfw text in my textview doesn't start rendering until I resize the window

ATaco

...feature?

Ginger

@cairdcoinheringaahing at least you didn't call me a sweet summer child d:

@ATaco not this time

caird coinheringaahing

@Ginger Oh you sweet summer child, I haven't even begun :P

Ginger

well now I've gone and done it, haven't I

Fixed it... by removing a call to set_default_size :|

ATaco

23:30

@cairdcoinheringaahing You whipper snappers don't know old.

lyxal

I know my grandma. She's hella old

Seggan

@RydwolfPrograms i mean any praclang that has a random interface could support it

for example, in java i could make a class extends Random and use ffi calls to TRNG methods or whatnot

ATaco

Lua and JS's randoms are explicitly PRNG. Though Node supports TRNG

user

@cairdcoinheringaahing *trying to be funny

Jacob

23:51

is trng true random

Transcript for