The Nineteenth Byte

lyxal

12:01 AM

@NoHaxJustRadvylf fine by me - I'll happily donate my share of message moving for the next few days to you

emanresu A

^

NoHaxJustRadvylf

That was several months ago :p

lyxal

Also why the username change?

NoHaxJustRadvylf

Life's too short for predictable usernames

lyxal

I know I haven't been with it for the last 2 days but there's nothing in the transcript about it :p

NoHaxJustRadvylf

12:03 AM

I just changed it like an hour ago

lyxal

Ah

Seggan

12:39 AM

@Steffan if it did I'd be blind from light mode

lyxal

I've had the best idea: twitter but you can only post AI generated content. All text is generated by a model trained on tweets with hate speech filtered out and all images are generated using things like craiyon. It'd be an interesting experiment

thejonymyster

@lyxal whats the user input

lyxal

5 words maximum + hashtags

Plus whether to have images or not

New Posts

12:59 AM

0

Q: Output your Code Golf submission from Try It Online

When using Try It Online, there is an option to copy a Code Golf submission: Using a language currently on Try It Online, your task is to output the exact submission copied to your clipboard from your code. For example, the following code in brainfuck: "Hello, World!"&@ Will create the followin...

OldSandboxPosts

Sandbox posts last active a week ago: Erverse Hte Ifrst Wto Eltters fo Aech Owrd, I X the source code, you Y the output!

Aiden Chow

2:12 AM

dang the guy from desmos discord posted the answer! codegolf.stackexchange.com/questions/249439/…

Actually insane

Sandbox Posts

6:05 AM

0

A: Sandbox for Proposed Challenges

In some languages, strings are started and ended with a quote mark ('). And quote itself is escaped by writing it twice sequentially. This question is about find out all strings in such format from the given input. Input / Output You are given a string contains only printable ASCII. Quotes (') in...

mousetail

6:39 AM

I should probably have added a minimum contrast requirement for the fractal challenge, it's so hard to tell if there are color differences

New Posts

8:13 AM

0

Q: Draw the USA flag

I noticed that it seems nobody has posted a non-ASCII (related) USA flag challenge yet, so what better time than the 4th of July? The Challenge Draw the current (as of 2022) flag of the United States of America. Here is the flag: Specifications Hoist (height) of the flag: A = 1.0 Fly (width) o...

code-golf kolmogorov-complexity graphical-output

Fatalize

@ais523 Are you still working on that Jelly/Brachylog hybrid?

ais523

9:05 AM

@Fatalize it's specced but probably too difficult to implement, so I put it on hold for now

Fatalize

Implementing a non-trivial language is always difficult

ais523

@PyGamer0 returning anything truthy would defeat the point, which is to continue processing from here only if all the elements are equal

@PyGamer0 it's a fail, not an error – this is a declarative language thing which effectively causes the code around it to fail until you reach something that can recover a failure, a bit like an exception

but the difference is that failure recoveries are really common in declarative languages, e.g. some sorts of loop will recover failure by omitting the iterations that failed, and anything that has alternatives will recover by trying a different alternative

I have been working on a golfing language that's more directly Jelly-inspired; there's some Brachylog influence there but it's clearly an imperative language

I added a null value which tends to propagate until you reach something that explicitly handles nulls, as a substitute for the fail behaviour in declarative languages

Unrelated String

smart

ais523

my main focus has been on the chaining rules, because Jelly's aren't very good in practice, frequently requiring extra bytes to disambiguate

the main trick is how nilads and dyads are handled; depending on context, either the encodings for dyads can be reused as secondary encodings for nilads, or vice versa (with the encodings for nilads reused as secondary encodings for dyads), or neither, depending on what's expected in the context

e.g. you very rarely want a random nilad in the middle of a chain (this triggers an "implicit print" in Jelly, which is only marginally useful behaviour), but it's very common to want two different chain shapes with dyads in

(in this case, they're "dyad whose argument is a single tacit variable" and "dyad whose argument is something else" – Jelly very frequently needs a ¹ or the like to distinguish these cases, so you're saving a byte every time that would happen)

here are the docs for what I have so far: Guide Encoding

note that these are unfinished and probably contain mistakes

(URLs should be fixed now)

I wasn't planning to release this publicly until I had something working, but it might be interesting to see where I've got to so far

emanresu A

9:33 AM

@ais523 Interesting place to host files :P

ais523

not really, it saves on domain names (I run the server)

emanresu A

Oh

@NewPosts We really need to figure out what to do with all these flag challenges :P

ais523

at some point you can start marking them as duplicates, but there's probably scope for a few interestingly different challenges along those lines

mousetail

I just collect random domain names I intend to re-sell at some point to host random stuff

Adám

9:47 AM

CMC Given a uniformly random ordering of 1000 unique numbers, what is the average longest streak of increasing numbers? Experimentally, I can determine it to be about 4.88, but is there a direct formula for that?

graffe

10:12 AM

hi all!

@Adám uniform from what range?

Adám

Does that matter?

graffe

there will be a lot of repetitions if the range is small

Adám

> unique

graffe

or are you sampling without replacement

ok

can you do it exactly for 2 and 3?

Adám

Well for 2 it'd be 0.5 because you either have 10 20 (max streak = 1) or 20 10 (max streak = 0).

graffe

10:17 AM

how can it be less than 1?

oh 20 10 is 0 and not 1?

emanresu A

Should be fine selecting a permutation of 1...n

Adám

@graffe Yes, I count the streaks as number of consecutive increases.

Unrelated String

^

graffe

@Adám ok

what is the answer for 3?

Adám

1

graffe

10:19 AM

actually, how about just counting how many for n =1 ... until you run out of time

and then putting it in oeis

Adám

1 2 3 : 2
1 3 2 : 1
2 1 3 : 1
2 3 1 : 1
3 1 2 : 1
3 2 1 : 0

graffe

try all permutations of length n for each n

@Adám nice

let's try 4

Adám

For 4, it is 1.291666667

graffe

can you show the breakdown?

Adám

1 2 3 4 : 3
1 2 4 3 : 2
1 3 2 4 : 1
1 3 4 2 : 2
1 4 2 3 : 1
1 4 3 2 : 1
2 1 3 4 : 2
2 1 4 3 : 1
2 3 1 4 : 1
2 3 4 1 : 2
2 4 1 3 : 1
2 4 3 1 : 1
3 1 2 4 : 2
3 1 4 2 : 1
3 2 1 4 : 1
3 2 4 1 : 1
3 4 1 2 : 1
3 4 2 1 : 1
4 1 2 3 : 2
4 1 3 2 : 1
4 2 1 3 : 1
4 2 3 1 : 1
4 3 1 2 : 1
4 3 2 1 : 0

graffe

10:21 AM

I feel there should be a recursive formula

that is to solve 4 in terms of 3

Adám

Ah yes, that makes sense.

The sequence is 0 0.5 1 1.291666667 1.491666667 1.638888889 1.766269841 1.875297619 1.971673832 2.056730875

graffe

to look it up we need an integer sequence

Adám

        0         1
        1         2
        1         1
161458283 124999961
 37291519  24999901
       59        36
  2596860   1470251
     6301      3360
   137124     69547
    93427     45425

No wait, that doesn't look right.

graffe

5

Q: longest consecutive subsequence of a random permutation

What is the expected length of the longest consecutive subsequence of a random permutation of the integers 1 to N? To be more precise we are looking for the longest string of consecutive integers in the permutation (disregarding where this string occurs). I believe the answer should be ~ c ln(n),...

co.combinatorics

that doesn't look very easy

I think it answers the question despite the question being unclear

Adám

Probably. Just was an interesting thing I stumbled upon when generating test data for a longest streak function.

@graffe Here:

     0     1
     1     2
     1     1
    31    24
   179   120
    59    36
  4451  2520
  6301  3360
141718 71877
 93427 45425

graffe

10:31 AM

what is that?

Adám

Enumerators and denominators of the results.

emanresu A

Here's something

Unrelated String

to characterize it as an integer sequence i'd think you'd want the numerators with denominators of, what, n^n?

versus reduced

er, n! probably

yeah

Adám

@emanresuA Agreed; that's something. But what is it?

graffe

can you compute it for n = 16?

Adám

10:36 AM

Me? No.

graffe

is that just 655536 checks?

ah no

oops

might be a fun challenge. I think there might be a clever way to do it

Adám

There are more than 10³⁵²⁸²⁹⁶ permutations…

graffe

isn't it 16! permutations?

PyGamer0

i got a idea that may not work for flax

graffe

20922789888000

@Adám why do you get such a big number?

Adám

10:41 AM

Oops, I did 655536! instead of 16!

But still, 20922789888000 is a lot of permutation to check.

PyGamer0

> You have already voted, but the voting has been cleared by a moderator

what

Adám

!

PyGamer0

i tried pinning a message in the flax room which was previously pinned

lyxal

How long ago was it first pinned?

PyGamer0

a month?

lyxal

10:44 AM

SE chat automatically unpinned it

PyGamer0

yeah i want to pin it again

lyxal

And you can't repin something automatically cleared I don't think

emanresu A

Just repost

Adám

8

Thank you, W⍺!

PyGamer0

lol

emanresu A

10:50 AM

Sep 5, 2021 at 11:43, by emanresu A

lyxal

Unrelated String

🤔🤔🤔🤔🤔

lyxal

@Adám hey at least it got the word right :p

graffe

@Adám yes. I suspect with some brain power it can be sped up

Unrelated String

that's honestly more ?????
because it indicates they actually mean to handle the specific phrase "deterministic permutation"
without considering what that would be meant to distinguish it from

PyGamer0

10:54 AM

((CMQ)): How useful are nulls in golfing?

also can anyone explain me what are choice points in brachylog

Fatalize

@PyGamer0 Choice points in Brachylog are the same as in Prolog and other declarative languages

PyGamer0

ah ... but what are they?

i do not understand logic programming

Fatalize

Let’s say you want a permutation of a string "test", in Brachylog you would write it as "test"p

PyGamer0

ok

Fatalize

But there are more than one way to satisfy this "declaratiion": as many as the number of permutations actually

PyGamer0

10:58 AM

@Fatalize ok so what do you mean by that?

Fatalize

So what Brachylog (Prolog under the hood) does, is that it picks the first one (based on how p was implemented), and keeps in mind all the other possibilities in mind, which are called choice points

@PyGamer0 "estt" would be a valid permutation, but so would "tset" or "test"

Adám

@PyGamer0 APL has nulls, and I barely ever use them. (I think I remember a single instance of using them.)

Unrelated String

with the propagation scheme ais mentioned earlier they could be reasonably useful

PyGamer0

like how would that work?

Unrelated String

but i feel like that might also be a bit of a weird compromise between declarative and not

PyGamer0

11:02 AM

@Fatalize so is this program saying true because test is a valid permutation of test? Try it online!

Fatalize

@PyGamer0 yes

Unrelated String

does have the advantage of, say, you can maybe manipulate the nulls as values in lists--maybe you have one builtin that sequences the list and turns it into a null if it has a null or leaves it alone if not

PyGamer0

@Fatalize oh ok .. how do i get all permutations thens?

Fatalize

@PyGamer0 "test"pᶠ

PyGamer0

does that assign all permutations of "test" to Z?

Fatalize

11:04 AM

What ᶠ does perfectly showcases choice points: Brachylog remembers that there are multiple possibilities as output for p, and ᶠ forces Brachylog to try them all and store the result

@PyGamer0 Here’s another example

Here we ask for a permutation which starts with "s"

Since p produces choice points, h"s" will fail for the permutations which don’t start with s until it tries a permutation that does

lyxal

So choice points are kind of like lazy evaluation that allow a sort of branching/tree exploration pattern?

Fatalize

pretty much

PyGamer0

@Fatalize makes sense

Fatalize

and Prolog/Brachylog are depth first

it tries all choice points of the latest predicate before it encountered a failure

PyGamer0

@ais523 so what would the failure handling thing do?

Fatalize

11:11 AM

@PyGamer0 I guess the idea is to have similar behaviour in an imperative language like Jelly

where instead of crashing when you input something that doesn’t fit, it tries something else until it fits

Ginger

11:23 AM

egg

mathcat

11:45 AM

Happy US-something day everyone!

Ginger

...

...correct

I find no problems with that statement

mathcat

Hey, I did draw the US flag on my exam paper because I had nothing to do!

Bgil Midol

ok

mathcat

or was it the USA flag

i don't care I'm not American

Ginger

the terms are interchangeable

mathcat

11:57 AM

gud

Seggan

2:24 PM

@mathcat Flag of the US

Happy Independence Day for y'all Americans!

mathcat

ah it was independence day

lyxal

Independence day? More like independence yesterday amiright? Ha gottem

Seggan

CMP: should implicit output be with or without newline?

lyxal

2:39 PM

With

Seggan

why?

user

You need some kind of separator

lyxal

And consistency with normal output

Seggan

@user why a separator at the end of execution

user

Why not? It'd look weird if you print out 1, 2, 3 as 123

Fatalize

2:45 PM

Without, you should output exactly what was asked to be outputted

lyxal

@Seggan wait are you talking about trailing newlines or something else?

Are you saying there could be more than one thing implicitly printed?

Because if so, then without newline between objects

Seggan

@lyxal trailing nl

@user its output as [1, 2, 3]

lyxal

@Seggan Well really it doesn't matter then

Seggan

@lyxal no

pxeger

I think there should always be a trailing newline

lyxal

2:54 PM

With trailing newline might be better for successive calls to the cli

pxeger

it makes almost no difference for use on TIO

but in a terminal, it can fuck up your shell prompt otherwise

user

^

mathcat

how come?

user

Your prompt continues right after the output

pxeger

bash $ python -c 'import sys; sys.stdout.write("hello")'
hellobash $

mathcat

2:58 PM

huh right

PyGamer0

@lyxal lol

mathcat

oh now I got it

Australian humor is strange /jk

user

Nah it's just lyxal who's strange

Adám

3:30 PM

ʞɾ/ ǝƃuɐɹʇs sı ɹoɯnɥ uɐıןɐɹʇsn∀

New Posts

0

Q: Erverse Hte Ifrst Wto Eltters fo Aech Owrd

Erverse Hte Ifrst Wto Eltters fo Aech Owrd Inspired by this accidental misspelling If you could read the title, then you already know what the challenge is about. If not, then I'll tell you. Easy: simply take each word (space delimited), and swap the positions of the first two letters. You may ...

code-golf

Adám

3:55 PM

Should we add an entry to

96

Q: Things to avoid when writing challenges

If there is something that you find annoying / counterproductive / unfair / detrimental / no longer funny in challenges (question posts), describe it in an answer here, and propose a recommended response. Voting will indicate whether that response has the backing of the community. For example, y...

discussion faq policy challenge-quality defaults

saying "posting, and then abandoning a challenge"?

I.e. "please post only when you're able to stay around for an hour to address immediate feedback."

PyGamer0

yeah i guess

user

It's definitely something that should be recommended to challenge posters, but it's not something you avoid while writing challenges

PyGamer0

^

Zion mycelia adamancy

4:20 PM

@pxeger What do you want to know about chain separators?

pxeger

I understand they separate chains. But I've not been able to get an intuition for what that means

The rest of the link after the separator is grouped into one chain?

or before the separator?

Zion mycelia adamancy

Basically, a chain is just a grouping of links. The first chain in the program is the links before the first separator. Then, each subsequent chain is the code between each separator, with the arity of the chain determined by the separator that starts the chain

e.g. AðBCµD is a link of 3 chains. The first is A and its arity depends on how many arguments are provided. The second is BC and is dyadic, and the third is D and is monadic

pxeger

oh, so AμBμC is not like A (B (C)), but (A, B, C), and when you write a link with no chain separator like ABC, it's actually one chain like (ABC)

Zion mycelia adamancy

Yes

pxeger

I always tried to mentally model them as nesting, like quicks do

Adám

4:30 PM

@user @PyGamer0 True, but even so, isn't it the most likely place for the advise to be spotted? Or do we have a guideline post?

pxeger

side note I discovered while writing that answer: you can rotate a list once to the left or right using ;Ḣ or Ṫ;

Zion mycelia adamancy

@pxeger There's also ṙ1 and ṙ- which don't modify the list

pxeger

@user @Adám This is advice that should apply to questions on every site on the network, so I'd say it has a broader scope than that question

I guess it probably applies more so here, because we tend to have a higher question validity floor

Adám

Also, I think people here hold themselves back from fixing questions for the first while, instead leaving constructive comments, and awaiting OP's fixes.

Ginger

4:51 PM

okay GB and OLIMAR should be back online

@Adám yes definitely (I'm pretty guilty of this)

Steffan

5:28 PM

What does it mean in Jelly if you "modify" a list? Does it seriously have mutable lists?

pxeger

It just uses Python lists, so yes

Steffan

That's so weird

Is the matrix tag appropriate for arbritrary dimension lists (not ragged)?

I think it's only for 2-D lists

If matrix is only for 2-D, maybe we should create a new tag

pxeger

6:01 PM

there's also grid, which could be multidimensional

although whether it makes sense depends on the task

I think we've had a previous discussion about what the difference between matrix and grid is

user

@Adám 🤷 There is that Welcome to CGCC post

Adám

> Some users may request additional clarifications in the comments. So long as you answer these promptly and don't contradict the main post, this won't be an issue.

Steffan

IMO, grid wouldn't be appropriate for multi-dimensional lists

Adám

OK, I've added it to the "Asking" answer there.

Steffan

A grid sounds like a grid, aka 2-D

A search for multi-dimensional array challenges doesn't find any tags but array

to me it would make sense to create a tag for multi-dimensional

user

6:20 PM

Btw @Steffan I just noticed you made a new language, is there a wiki for it yet?

Steffan

nope

user

Is it going to be like Jelly?

Steffan

6:33 PM

No, more like Fig and Pip

user

Cool

Seggan

Yikes, I post a challenge, go to the lake, and 2 hours later I get 20 items in my inbox??

@Adám yeah sorry I didn’t mean to leave so quickly

Thought I had an extra half hour before going to the lake

Adám

No worries.

Seggan

CMP: y’all want a Fig parsing challenge, my math shorthand parsing challenge, or a k means challenge?

Note that I haven’t checked for dupes yet

Ah k means exists

mathcat

6:53 PM

Quick poll, can you explain on the spot how neural networks work?

hyper-neutrino

not sure if this is accurate, but from my understanding, you have a collection of nodes that are like neurons in layers where each layer has connections between from one's nodes to the next with the first being the input and the last being the output, and each node has some way to calculate its individual output given its inputs, and then the input and output is given and received by presenting the data to the network as a set of weights

mathcat

lol I expected yes/no answers

Seggan

Iirc each weight is basically how much impact that neuron has on the next one

There was a veritasium video that explained it really well

mathcat

nah 3b1b's video was better

@hyper-neutrino Can I expect the average programmer to know this?

pxeger

That's just the very very basics, so you can certainly expect the average programmer to understand it given a single paragraph of explanation

but I think it's not common enough knowledge to rely on most people already knowing it unless they've done some AI before

hyper-neutrino

7:04 PM

I would more so question if it's useful to expect the average programmer to know this

like, I've known that the derivative of x^2 is 2x for a bit longer than I've understood how calculus works and it was not useful to know that because I had no basic understanding - knowing that neural networks can be represented as a combination of consecutive matrix-vector products and convolution functions (IIRC that's how it works?) does not really tell me anything about how they work lol

mathcat

It's for a presentation about something that has to do with AI, I was thinking of leaving the explanation out

hyper-neutrino

oh, I see

Steffan

7:42 PM

@user There is a horrible Scala.js online interpreter, and a list of functions here. It's just prefix functions like Pip and Pyth and Fig

NoHaxJustRadvylf

@Seggan I don't think I've ever heard someone call it that before lol

Other than calendars

Steffan

Yeah I just call it 4th of July

user

@Steffan Ooh nice

NoHaxJustRadvylf

It's really weird how everyone just calls it by the date it occurs on. Not sure if any other holidays are like that

Steffan

I think there is at least one other

Like a Mexican one

NoHaxJustRadvylf

7:43 PM

Oh yeah

Steffan

Like 24th of May or something, idr

NoHaxJustRadvylf

Cinco de Mayo

user

Nah that one's named after a condiment

NoHaxJustRadvylf

May fifth

Steffan

Oh yes

user

7:44 PM

Yeah, it's on May 5th that mayo sank

3

Steffan

Bruh why couldn't I remember that

user

It sank from heaven and fell to the Earth then, which is why it's celebrated then

Steffan

We should call today Cuatro de Julio instead

Because Cinco de Mayo is usually more interest in the US than Mexico even

user

I don't think we need any more interest in Independence Day

I haven't gone out today, but in previous years everything's red, blue, and white on July 4

Steffan

lol

NoHaxJustRadvylf

7:47 PM

Patriotism is for nerds

America sure does love itself for being a somewhat below-average (western) country

pxeger

CMM: if a challenge explicitly bans output to stderr, but allows all other standard I/O formats, can I output to a file which just happens to be called /proc/self/fd/2?

emanresu A

git commit utter frick

NoHaxJustRadvylf

@pxeger I would say yes technically, since there's probably some system out there where that could be an ordinary file, but I'd probably downvote

I don't know why people get all mad when a challenge overrides standard I/O, it's literally only there as a default to fall back to

hyper-neutrino

I suppose Juneteenth is sort of named after the day on which it occurs.

pxeger

(I think the answer is a pretty clear no. /proc/self/fd/2 and stderr are, by definition, identical. I just thought it was funny)

NoHaxJustRadvylf

7:58 PM

And overly permissive I/O is incredibly boring, and our standard I/O has to be overly permissive

@hyper-neutrino That one annoys me since I never can remember which of the seven days ending in "teenth" it is

hyper-neutrino

well, the N in June makes it sound closest to nineteenth compared to the other teen numbers (and also nineteenth and Juneteenth share the longest suffix of the seven possibilities) :P

NoHaxJustRadvylf

Yeah, that makes sense, although I still for some reason always end up thinking it's the thirteenth (I guess because of the thirteenth amendment) lol

emanresu A

Speaking of which

-1

A: Default for Code Golf: Input/Output methods

Programs may use the camera and microphone I wouldn't expect to see this very often, but it could be applicable to my upcoming challenge.

NoHaxJustRadvylf

That's...really not what standard I/O is meant for lol

> it could be applicable to my upcoming challenge

Then add it to that one :p

I honestly think our standard I/O is a huge mess and should be rewritten from the ground up but that's just not really worth it

pxeger

I maintain that I/O through predefined variables should have been allowed

NoHaxJustRadvylf

8:06 PM

I don't think it should, but that causes problems for esolangs

Having one set of rules for all types of languages causes a lot of issues

Honestly I think language creators deciding how their language is allowed to do I/O, and us deciding per-language or per-class-of-similar-languages for ones not made by CGCCers, makes more sense, but it's a lot of work

E.g., doesn't matter how cheaty a language's I/O should be, just like with flags, and we just need the same sort of "go too far and we'll downvote you" attitude

hyper-neutrino

honestly it could make sense to just not have these set requirements

so many exceptions exist for languages that require them like /// not even being able to take I/O except through direct insertion that it's hard to really standardize

NoHaxJustRadvylf

I think it'd make sense to have a set of standard I/O guidelines, maybe with different sections like "practical languages with lambdas", "practical languages with STDIO", "turing tarpits", "golfing languages", etc., and allow challenges or submissions to break those suggestions if they feel it necessary

emanresu A

Cursed idea: land where syntax errors can be wrapped in Try/catch

NoHaxJustRadvylf

We could point new users toward those guidelines, and strongly recommend that they, e.g., use a function instead of outputting via a variable, and if a user goes too far with the I/O we just downvote or use the classic "taking additional input" or "producing additional output" loopholes

pxeger

@emanresuA You will shortly be arriving at zsh land. Please mind the gap between your brain and the platform.

6

);echo hi

or: codegolf.stackexchange.com/a/231542

emanresu A

8:14 PM

Exceptions are for exceptional cases. Otherwise, “one I/O to rule them all” just works.

NoHaxJustRadvylf

We have a lot of exceptional cases though, and there is no default language

Half the praclangs have lambdas and half don't
Half the praclangs have access to STDIO and half don't (e.g., non-Node JS)
Half the languages used here are golfing languages with implicit I/O
Some esolangs lack any I/O at all, and many are far from exceptional

pxeger

@emanresuA I just noticed the typo "land" here. I unintentionally punned on it in my reply lol

Presuming you meant "lang", and not that your compiler changes core behaviour when crossing international borders

emanresu A

@pxeger Terrible idea #2!

pxeger

@user An cat called "one two three" and a cat called "un deux trois" decided to have a swimming race across the English Channel. The English cat won, because un deux trois quatre cinq

emanresu A

@NoHaxJustRadvylf with the whole suit based on objectivity, having subjective validity Croats

pxeger

8:20 PM

(but what if the six sept?)

emanresu A

@NoHaxJustRadvylf with the whole suit based on objectivity, having subjective validity criteria is kinda counterproductive.

NoHaxJustRadvylf

I'm doing a flag rule here and making it perfectly objective: Any I/O is valid

If you do something stupid tho, you get downvoted and probably violate a loophole

emanresu A

@NoHaxJustRadvylf this was what the early site was like and it was chaos

NoHaxJustRadvylf

E.g., in JS, using x=>y=> for input to save a byte is pretty normal, and since currying inputs is allowed (presumably for languages like Haskell and...functional tarpits? those a thing? stuff like SKI), nobody really questions it. But I find it an incredibly boring byte save.

@emanresuA We've matured quite a bit. The flag rule has been pretty successful at keeping things like MGS at bay, so I have no doubt this would for I/O stuff

And I don't know about you, but I don't really look at standard I/O in the first place. If it seems reasonable to me, I assume it's allowed, and if it seems unreasonable to me, I assume it isn't.

pxeger

It would be good if every language had exactly one standard IO method. That would enforce the "keep your golfing in your code, not your IO" rule

NoHaxJustRadvylf

8:24 PM

@NoHaxJustRadvylf ^ That's sort of what I'm aiming for with this

emanresu A

By doing this, 90% of submissions will become uncompetitive. It’s that really whaa a t you want?

NoHaxJustRadvylf

I'm not sure what you mean

Uncompetitive as in not very golfy, or uncompetitive as in breaking rules?

My goal is to decrease the cheaty I/O, not increase it

emanresu A

As in those answers could easily save some bytes by doing variable I/O or similar

NoHaxJustRadvylf

Having a site-wide rule saying "xyz is allowed" encourages people to apply that rule where it doesn't belong. You look at the 500 allowed input methods and pick whichever one is golfiest in your specific situation, and optimize there instead of making your actual submission interesting. "Anything goes" means that when someone suggests "hey, you can save a byte by taking input with a variable", you have the power to say "no, that's an unreasonable way to take input, and it's really nonstandard"

@emanresuA And variable I/O is pretty cheaty

Just like how saving bytes with a ton of flags is

emanresu A

I’m fat, you don’t need variables at all

NoHaxJustRadvylf

8:29 PM

If the guidelines say "taking input with variables is strongly recommended against in languages with functions", then if you take input with variables in a language with functions, you'd better have a good reason

emanresu A

You can just insert input into the code

NoHaxJustRadvylf

@emanresuA Was that supposed to be "in fact" lol

emanresu A

Oops

NoHaxJustRadvylf

Mobile moment

pxeger

My ultimate goal for the rules would be to never have to think about it at all. I think variable IO is the best way to achieve that

NoHaxJustRadvylf

8:29 PM

@emanresuA Which is even cheatier and more nonstandard, and actually allowed by our current standard I/O in some cases

emanresu A

Only in exceptional cases

And how do you define what’s cheating and what’s not?

NoHaxJustRadvylf

And there's no reason that it would be any less taboo without hard limits on what you can do

@emanresuA I define it how I want to define it and vote accordingly

That's the power of subjectivity and the beauty of the flag law

emanresu A

10 mins ago, by emanresu A

@NoHaxJustRadvylf with the whole suit based on objectivity, having subjective validity criteria is kinda counterproductive.

NoHaxJustRadvylf

I disagree, it's counterproductive having to explicitly ask permission every time you want to use a language that doesn't conform to our standard I/O, or spend time golfing your I/O instead of the fun part

This moves nonstandard I/O from illegal dark magic to a tool you have for edge cases, and one that, if misused, will only hurt you

I gotta go for now but this is an interesting discussion

Zion mycelia adamancy

@pxeger I dislike this: if a Jelly program has more than 2 inputs, it's genuinely part of the golfing to change the I/O method

Unrelated String

8:55 PM

i like it as an aesthetic concept but yeah absolutely allowing multiple i/o conventions is healthy for adding further depth to golfing

thejonymyster

9:07 PM

what are we discussing where do i start backreading

Ginger

9:18 PM

^

user

Perhaps around here or further back, when New Posts announced Seggan's question

emanresu A

9:33 PM

Jan 28, 2019 at 20:29, by New Main Posts

DON'T WORRY. WRONG NUMBER

thejonymyster

ah i see, interesting fiscusdion

i actually suggested something like the "use hardcoded input in langs with regular i/o" and it got downvoted

Zion mycelia adamancy

@thejonymyster fiscal discussion? :P

thejonymyster

oops i missed

i agree with the sentiment that we shouldnt have to really check all the time for what should/shouldnt be allowed, but at the same time its nice to link to something that sort of shows "this is what the community generally feels about this"

instead of having to argue its reasonability on any given post

personally i dont even like deciding what sorts of i/o are cheaty on my own questions. i feel like if the question is specified enough, it should be the community's decision whether something is cheap or not

i guess mostly it depends on it something is questionable in terms of "the spirit of the question"

thats my... words i will say

my "take" if u will

emanresu A

10:47 PM

CMC: Given a list of integers I with length l, consider a list of variables with the same length such that v_1 + v_2 + ... + v_n-1 + v_n+1 + v_n+2 ... + v_l. Your challenge is to calculate the values of the variables.

For example, with a list of length 3 e.g 8, 6, 4, the equations represented are y+z=8, x+z=6, x+y=4. The unique solution is z=5, y=3, x=1, so the output should be [1, 3, 5].

Neil

@NewPosts surprised how easy this was without complex arithmetic

Adám

@emanresuA APL, 7 bytes: ⊢⌹⍋∘.≠⍋

emanresu A

11:10 PM

How does that work?

I recognise a gradeup

Bubbler

Create the matrix that represents the system of equations and "matrix divide" the input vector by it to solve the equations

emanresu A

Huh, cool

Adám

11:32 PM

@emanresuA ⍋∘.≠⍋ is just a golfing trick to generate a NOT'ed identity matrix. Each row in it represents 1 of each variable, except for one (on the diagonal). And then, as Bubbler writes, ⌹ solves the equation system with the given sums (⊢).

Neil

11:45 PM

@emanresuA pretty sure it's just something like sum(I)/(l-1)-I

Transcript for