The Nineteenth Byte

Zacharý

12:00 AM

That'll work, I think: tested with -

Adám

@EriktheOutgolfer Why? (Other than you may want ⍪⍀ for higher-rank arrays.)

Zacharý

12:32 AM

I have the potential row/column generating function of my solver... but I realized I made a horrible looking dfn ... >_<

Adám

@Zacharý Sometimes it better to go with a tradfn.

Zacharý

I'm going to rewrite it as a tradfn, because this:

{0<T/⍨{(∊⍳¨W)≡⍵~0}¨T←(+×⊢){⍺⍺⍨/∘⌽¨⍪\⍵}¨{1@⍵⊢A⍴0}¨S/⍨{∧/(⍵>A)⍱1>⍵}¨S←∪¨∊¨(1-⍨⍳¨⍵‌)∘+¨S/⍨</¨S←⊃{,⍺∘.,⍵}/(≢W←⍵)/⊂⍳A←⍺}

is NOT good coding style.

Pavel

oof

Zacharý

^, exactly

Pavel

I'm a scrub so I define tradfns for everything with more than 2 bytes

Adám

12:37 AM

@Zacharý Wait a minute, it doesn't look so bad, you just need to ungolf it: Insert a line break at all inline assignments, and avoid parens by using assignments for sub-expressions and functions.

H.PWiz

I think I have written about 2 tradfns

Zacharý

^ other than the student competition, that's exactly the same as me.

One in GNU APL, the other in Dyalog (a failed attempt at what I'm doing now)

Adám

Maybe we should do one or more in-depth lessons on coding style? Maybe one for tradfns, one for dfns, and one for tacit fns?

Zacharý

Wait a minute ... there's STYLE for tacit fns?!!

I'll think I'll write it in a tradfn, as I'm going to be doing stuff after the generation function...

Adám

@Zacharý Another thing you need to do is use descriptive variable and function names. And maybe add some comments. I couldn't at a glance find good names for everything, but here's a start:

 f←{
     A←⍺
     W←⍵
     Wlen←≢W
     AllPairs←,∘.,
     AllCombos←⊃AllPairs/
     S←AllCombos Wlen/⊂⍳A
     inds←1-⍨⍳¨⍵
     AllInds←{∪∊inds+⍵}
     S2←AllInds¨S/⍨</¨S
     InRange←{∧/(⍵>A)⍱1>⍵}
     S3←S2/⍨InRange¨S2
     Put1s←{1@⍵⊢A⍴0}
     BackScan←{⍺⍺⍨/∘⌽¨⍪\⍵}
     T←(+×⊢)BackScan∘Put1s¨S3
     No0s←{(∊⍳¨W)≡⍵~0}
     0<T/⍨No0s¨T
 }

@Zacharý John Scholes (the dfn guy) is a big proponent of "denotative programming". He says that you can get better code just by reading ← as "denotes". This way you'll have to use meaningful names, and you can't reuse names.

@Zacharý Sure. And with version 17.0 you'll get some tools that make hard-core tacit programming much easier.

Zacharý

12:52 AM

What? I can't think of anything like that off of the top of my head, guess I'll wait till v17.0... when is that planned to be released? Or is there not a firm deadline?

Adám

@Zacharý Not a firm deadline, but we're aiming for end of Q2. Code freeze is in about a month. Then we'll start the beta programme and focus on documentation and bug-fixing.

Zacharý

Woah: I'm gonna be aging about the same time (give or take a month or two) as Dyalog's version numbers... :)

Adám

@Zacharý I plan to be using it in next week's webinar. It isn't the subject, but you can still see it in action if you pay attention.

Zacharý

Will the webinar be recorded?

Adám

@Zacharý Of course. Like always.

Zacharý

12:57 AM

Good...

Pavel

@Adám Open beta or closed beta?

Adám

@Pavel Closed. Mainly for paying customers to see if their existing systems will run on the new version, and maybe to compare speed too. We're expecting some pretty significant speed-ups. E.g. today, Marshall implemented an idea of mine that makes monadic ⎕UCS 10–50 times faster.

Pavel

Aww. Where does that put TIO?

Zacharý

Don't push it... TIO got the license for free, IIRC.

Adám

@Pavel TIO should probably wait for a stable release, since it doesn't have the functionality to test existing code. If something is wrong, all affected TIO links will become non-functional with no alternative.

Speaking of which, I should probably begin formulating my fastest-code challenge…

Pavel

1:06 AM

@Adám The different versions are installed in different directories anyway iirc, so it would be easy to roll back

Adám

@Pavel I don't think so. When Dennis pulls a repo, the old version is gone (from the online TIO anyway), no?

Pavel

Yeah, but he could just reinstall from the old rpm

Adám

@Pavel I'm not going to bother him with that. TIO can wait. TryAPL is still on 14.0! sigh.

Pavel

D:

Adám

@Pavel It is almost ready for upgrading, there's just one tiny issue with launching APL processes on Linux. It all works under Windows.

Οurous

1:14 AM

@Adám <drools at fastest-code tag>

Zacharý

Speaking of challenges, I might make a few related to <insert-preferred-name-of-puzzle-i-am-in-the-process-of-solving> ( :p ).

Οurous

@Zacharý Minesweeper! Chess! Tetris! Checkers! Elite:Dangerous! Factorio! :P

Zacharý

Meant puzzle, sorry

Solve Chess? Are you high?!

Adám

Clone KSP code-golf /s

Zacharý

What is KSP?

quartata

1:19 AM

kerbal space program

Zacharý

ah... still don't get the joke

Οurous

@Zacharý I prefer "optimistic" to "high"

Zacharý

@Οurous So ... HIGHly optimistic then?

Adám

@Zacharý KSP is a complex physics heavy game with a lot of data and GUI.

Zacharý

Ah.

Can tradfns modify their arguments (I don't think so, just checking)?

Adám

1:23 AM

@Zacharý Yes, for sure:

ATaco

@Adám Bash, cp -R ksp ksp2

Adám

num←Double num
num×←2

Zacharý

Ah ... I'll just store it in a temporary variable for the sake of ⍣ then.

Adám

@Zacharý Tradfns have quite a few tricks that dfns just can't do.

Zacharý

Returning a function is one of them, right?

Adám

1:29 AM

@Zacharý Yes, but also be niladic or strictly monadic or dyadic, put restrictions on their right argument, dynamically shadow names, use user defined functions for modified assignments, use multi-condition control structures, handle concurrency with tokens (semaphores),…

Zacharý

I would think using user defined functions for modified assignments would work in a dfn...

Adám

@Zacharý It won't, because dfns auto-shadow any names to the left of a ←.

Zacharý

Ah

Adám

@Zacharý You can use inline dfns and tacit fns in modified assignments, but if you try to give them a name first, it won't work anymore.

@Zacharý Actually, I just now found an ugly hack to do it in a dfn.

Zacharý

^ what is it

Adám

1:42 AM

@Zacharý {num←⍵ ⋄ f←× ⋄ num (⍎'f')←2 ⋄ num}

Zacharý

Doesn't look TOO bad...

New Main Posts

0

Q: Displaying Tweets

This challenge is about displaying user's tweets. Write two versions of program that prompts user to input tweet, the format and display it. (one with 140 characters long limit and one with 280 characters long limit) Rules: Formatting: 1. If there's any usernames (beginning with @ character) ...

ATaco

@Adám Your comment may be considered hostile based on how it's worded, but I agree with you.

Adám

@ATaco Thanks, I wanted to comment quickly. Better now?

ngn

@Adám also: num f∘⊢←2

Adám

1:51 AM

@ngn Ah, right. Nice. That's acceptable (as opposed to the ⍎ hack.)

Zacharý

@ngn that actually looks good (as in not an "ugly hack")

ATaco

Works a lot better.

Adám

@ngn also: (⊢f)←2

ASCII-only

Hmm I wonder if Zachlike tips questions would be on-topic here

Adám

@ASCII-only Zachlike?

Zacharý

2:03 AM

@ASCII-only ?

Zacharý

2:14 AM

Can you please explain?

ASCII-only

@Adám e.g. Shenzhen I/O, TIS-100, Silicon Zeroes

They're kinda like visual programming languages

Adám

@ASCII-only That did not clarify anything for me. Sorry. I'm probably just dumb.

ASCII-only

@Adám They're games made (mostly) by Zachtronics

Nick Clifford

Broccoli update! after lots of work, the interpreter can now fully interpret a basic test program! all that's left to do is implement the rest of the builtin functions originally described by the original author. I am looking for some people to help with this, so if you know any C# and are interested in helping, please join the PPCG Podcast Discord server (discord.gg/9n4dAzH). more details there. thanks!

ASCII-only

@NickClifford :/

Adám

2:24 AM

@ASCII-only Ah, now I get it. Those games have you write "programs" that need to be as short as possible. You're wondering if they count sufficiently as programming languages such that tips for golfing in them are on-topic, right?

@NickClifford Is Broccoli a Lisp variant?

Nick Clifford

@Adám it's a hellish combination of functional and imperative programming styles that happens to look like Lisp.

DJ can attest to that

ASCII-only

@NickClifford hmm I should install VS first right

Nick Clifford

@ASCII-only assuming you're on windows, yeah

Zacharý

@NickClifford So ... lisp + APL? (With the functional + imperative, at least)

ASCII-only

time to use up a lot of my uni internet quota

Adám

2:29 AM

@Zacharý kparc.com/lisp.txt

Nick Clifford

@Zacharý moreso lisp than anything. hold on, let me find a larger example program

here's a primality tester written by DJ after we discovered Broccoli on the first episode of the PPCG podcast

(fn p ($a)
    (:= $d 0)
    (for $i in (range 2 $a)
        (if (= $a (* $i (int (/ $a $i))))
            (:= $d (+ $d 1))
        )
    )
    (= $d 1)
)
(map p (range 1 20))

much better

Zacharý

... no wonder this is called Broccoli: some people might love it, while others just want to vomit.

5

Nick Clifford

here's a quote from the original README

> Broccoli is an exciting new programming language that is guaranteed to make you pee your
pants (in a good way).

Zacharý

... burn it with fire

Pavel

TIO when?

ASCII-only

2:38 AM

When it's finished probably?

Nick Clifford

@Pavel we actually tried to get the original Windows executable on TIO, but Dennis couldn't get Wine working

Pavel

@NickClifford Didn't you just say the C# version works now

Zacharý

@flawr Wouldn't it be 528,000 centipedes in a mile?

Nick Clifford

the test program works. it's not up to spec yet though

Pavel

@NickClifford Well that's fine, it can always be pulled as more stuff gets added

Nick Clifford

2:42 AM

true, true

Pavel

(I'd ask Dennis myself but I forgot how to HW)

ASCII-only

I'd wait

hopefully it will be more complete by the end of today

hopefully

Nick Clifford

ayy

that's the spirit

Zacharý

2:55 AM

@Pavel What does "HW" mean in this context?

Pavel

hello world

Zacharý

Ah

Take "not up to spec yet" w/ a grain (or a mountaint) of salt. the test program could be the most basic thing ever.

Downgoat

4:21 AM

is it just me or do other people here type '1' in the print dialogue expecting it to print the second page?

Zacharý

@Downgoat ?!

DJMcMayhem

@Zacharý @NickClifford The best way I can think to describe it is if you had lisp but didn't allow functions to be first class citizens

Zacharý

@DJMcMayhem ... sounds APL-ish

DJMcMayhem

I know very little APL, but from what I do know, it's about as far away from APL as you can get

Zacharý

The only thing that makes it similar is its non-functional functionalness

Pavel

4:28 AM

It's an imperitive language, that happens to look like a Lisp (but really isn't like normal Lisps), with some functional builtins

Like how python has map and filter but you wouldn't call it functional

Zacharý

Or its non-imperative imperativeness

Downgoat

@ASCII-only whats our slice syntax

and should slice return a new array or a ref to a subsection?

Οurous

4:57 AM

@Downgoat I'd think a new array if they behave like by-value lists, or a ref to a subsection if they're like C pointer arrays

Downgoat

@Οurous they are more like a Java List than a C-array

Pavel

@Downgoat If they're like Java ArrayLists then a slice should return a new array.

If they're like Java arrays, you shouldn't ever be able to ref into the middle of one

Calvin's Hobbies

Just popping in to say rest in peace Stephen Hawking :/

quartata

yeah I'm really bummed out

ASCII-only

@Downgoat :/ did we have one

@Downgoat I'd say Python's isn't a bad idea in this case

Downgoat

5:05 AM

@Calvin'sHobbies :( Stephen hawking had this sort of bad-ass physics vibe to me that I can't really explain. I still don't understand any bit of M-theory but sad day to see a great scientist go. :(

7

@ASCII-only idk

ASCII-only

5:32 AM

anyone know what to call a function that takes non-evaluated arguments?

like if if was a function

Downgoat

@ASCII-only macro

Esolanging Fruit

@ASCII-only Normal, if you're using Haskell.

Otherwise, probably control-flow macro.

Downgoat

5:56 AM

btw @ASCII-only idk if I said but lazy vars work now :DDDD

; ModuleID = 'main'
source_filename = "main"
target datalayout = "e-m:o-i64:64-f80:128-n8:16:32:64-S128"

%0 = type { i32, void ()*, i8* }

@"x.$307.value" = private global <{ i1, i32 }> zeroinitializer
@0 = private unnamed_addr constant [10 x i8] c"got int32\00"
@llvm.global_ctors = appending global [1 x %0] zeroinitializer

define internal i32 @"x.$307"() {
entry:
  %0 = load i1, i1* getelementptr inbounds (<{ i1, i32 }>, <{ i1, i32 }>* @"x.$307.value", i32 0, i32 0)
  %1 = icmp eq i1 %0, false

is what you get for:

lazy let x: Int = 1

func main() {
    print(x)
}

not sure how to reduce it to 1 load

but it should hit CPU cache pretty well

additionally branch weights should be correct so branch prediction should be on our side

@Pavel This will throw a syntax error in VSL beacuse anyone who does that does not deserve working code

Manish Kundu

codegolf.stackexchange.com/questions/158060/which-card-is-it is it okay to tag this as kolmogorov complexity?

New Main Posts

0

Q: Which Card is it?

Introduction Long ago, when I used to code card games with usual playing cards, I used to specify a number for each card and call a function with some number to get a card. This somewhat inspired me to make this challenge. So for people unaware of the playing cards, a deck of cards consist of ...

code-golf string kolmogorov-complexity integer

xnor

6:12 AM

@ManishKundu this is very cheesy, but choosing any start number could let the golfer encode their program in the lows bit of every input

Manish Kundu

so just 1 and 0 as the start number should be done?

xnor

i think so, I don't see anything interesting people could do with other values that isn't an exploit

Manish Kundu

alright.

New Main Posts

6:58 AM

-1

Q: Print all the positive non-zero integer n-tuples that solve an inequality within a bound

Suppose I have a linear inequality like x_0A_0+x_1A_1+...+x_nA_n <= C With x_i a non-zero positive integer and A_i & C a positive non-zero multiple of 0.01. Find all the positive integer n-tuples {x_i,...,x_n},{y_i,...,y_n},...,{z_i,...,z_n} such that: D <= x_0A_0+x_1A_1+...+x_nA_n <= C Where...

code-golf math code-challenge

user202729

8:31 AM

Why are the (mod) diamonds red for me?

betseg

Your font?

betseg

8:51 AM

@user202729 ?

user202729

Anyome know of a keyboard on android that can type all Jelly characters?

Manish Kundu

9:35 AM

probably not. maybe create a challenge asking people to make one :P

ASCII-only

9:51 AM

@user202729 American

i.e. English (US)

Erik the Outgolfer

12:34 PM

@Adám who's to say that the array only contains simple scalars?

Adám

@EriktheOutgolfer Should still work. Counterexample?

Erik the Outgolfer

@Adám (1 2 3)(4 5 6), while (⍳∘≢↑¨⊂)∘⍉ returns correctly, ,\ doesn't

Fatalize

@EriktheOutgolfer Currently learning a few greek words: the pronounciation of the letters is a mess!

why the hell does μπ make a /b/ sound?!

Erik the Outgolfer

well, it might be that people being bored mispronounced it and it eventually became standard :P

Adám

@Fatalize 0.000003141592654

Fatalize

12:40 PM

also why is there ten billion different spellings for the sound /i/?

@Adám π_π

Erik the Outgolfer

@Adám that...isn't exactly how μ works

Martin Ender

@Fatalize ...says the French guy ;)

Fatalize

also why did they chose "yes" to start with a /n/ and "no" to start with a /o/, there's no way I'm not gonna confuse the two at least once

Erik the Outgolfer

^^ exactly that, even English is messier than Greek!

Adám

@MartinEnder …says the German guy ;)

Erik the Outgolfer

12:43 PM

uh, I don't think that's the kind of reply that can be chained like that

Adám

@EriktheOutgolfer But Danish is the messiest of them all!

Fatalize

German pronnounciation with regards to spelling is easy as hell

Erik the Outgolfer

@Fatalize η ι υ ει οι υι, but some of them are actually supposed to be /i:/

(even though that was left far behind when Ancient Greek was replaced by Koine Greek)

Adám

E.g. in Danish, the vowel "u" has 20 official pronunciations — and that's just when it isn't affected by neighbouring letters.

Fatalize

@EriktheOutgolfer Can you get by in English in Athens?

Giuseppe

12:47 PM

@Fatalize I did OK there for three days doing tourist-y things (I was on a study abroad trip)

Erik the Outgolfer

on tourist areas, pretty much yes, but on more local areas I'm afraid...

(unfortunately, the tourist areas are the expensive areas)

Fatalize

I'm only here for a conference near the center of Athens, so I guess I should be ok

I hope they speak English at MacDonald's so that I actually survive

Erik the Outgolfer

basically it's older people who mostly have issues with English

you mean at Syntagma square? I'm pretty sure you won't have any problem there

Fatalize

Close to it

Closer to the byzantine museum and war museum

Erik the Outgolfer

yeah I know what you mean, although <100m is way closer than just "close" :P

J. Sallé

1:00 PM

@Fatalize Greek food is delicious though, you should try some.

Fatalize

@J.Sallé I hate eating and food in general, I only eat to survive

8

J. Sallé

Fair enough

NieDzejkob

@Fatalize and yet you choose McDonalds...

Erik the Outgolfer

maybe since he's used to it

Fatalize

@NieDzejkob Well yes, because it's easy to eat

and I know that the one in Greece will be roughly the same as the one I'm used to

NieDzejkob

1:03 PM

if you eat to survive, eating at McDonalds is a bit... counterproductive

Erik the Outgolfer

although if you don't try pork gyros souvlaki with tzatziki you are committing a crime :P

Fatalize

to survive on a short time-scale

J. Sallé

@EriktheOutgolfer tzatziki is the yogurt sauce isn't it?

Erik the Outgolfer

well, basically yes, but in fact it tastes very different

J. Sallé

@EriktheOutgolfer yeah I've had it once at a friend's restaurant. Tastes good with those bbq'd skewers

Erik the Outgolfer

1:15 PM

lol in Greece that's kalamaki, not skewers :P

J. Sallé

Hahahah I'm pretty sure there are correct names for everything, I just don't know them. The eating part is what I'm really good at :p

Erik the Outgolfer

no, kalamaki isn't skewers

it's literally a wooden stick, pointy on one end

ASCII-only

@EriktheOutgolfer So basically skewer sticks

Erik the Outgolfer

yeah, the sticks, not the metallic ones

ASCII-only

what is this heresy

people use metallic skewers :O

Erik the Outgolfer

1:28 PM

uh

J. Sallé

Aren't wooden skewers also called skewers though? The name is the same in Portuguese >.>

Erik the Outgolfer

well, yes, but metal skewers aren't called kalamaki

J. Sallé

Ah, I get it now. Translation issues

ngn

2:04 PM

@Fatalize because "b" is a voiced bilabial stop - "m" gives the voicing and "p" gives the stop, if you try to pronounce them simultaneously :)

Erik the Outgolfer

1 hour ago, by Erik the Outgolfer

well, it might be that people being bored mispronounced it and it eventually became standard :P

Manish Kundu

So only 3 uncracked entries in cops vs robbers challenge?

btw how do you write hyperlinks in comments?

NieDzejkob

@ManishKundu [text](url) -> text

Manish Kundu

ty

J. Sallé

3:10 PM

Hey friends, any more feedback on this would be appreciated.

user202729

3:21 PM

@J.Sallé I'm afraid that any changes to the normal scoring system can be exploited, unless you're very careful.

(the "b and # counted as 1 bytes each" part)

Zacharý

Huh? Who's talking about Modern Greek?

J. Sallé

@user202729 I see. I added that because I believe everyone would have to use, at least, the string version of flat and sharp to output it

Also, I've never seen a lang use them as code, but then again there are thousands of answers I haven't seen here :p

user202729

3:51 PM

Can anyone check if this is selfcontained enough?

(probably not, but do you have any suggestion to makeit better?)

Zacharý

I think I found my problem in my code for Jagony.

Whenever I call ".Chain", it references and calls upon something that isn't defined yet, namely " System.MainClass.DeclarePublicStaticMethod", so it defaults to the blank object... I have no idea how to fix this

ngn

@user202729 "(n-1)%(l-1)+1th link (1-indexing)" - 1-indexing is distracting and only forces you to add +1

Zacharý

4:09 PM

In APL, would changing the naming mechanism a bit to where only names beginning with a ... z ∆ are functions, and A ... Z ⍙ being operators make APL parseable? Or do monadic/dyadic operators have to be split up further?

Erik the Outgolfer

> (n-1)%(l-1)+1th link (1-indexing)

shouldn't it be "n%lth link"?

ngn

@Zacharý to make apl parsable, you need to know the "kind" of every name in every scope: (a.) array / niladic function, (b.) monadic/dyadic function, (c.) monadic operator, (d.) dyadic operator

Lembik

basic question but.. when you normalize an array to have unit length and mean zero

do you subtract the mean after you have made it unit length or before?

cs.stackexchange.com/a/88864/10359 suggests before

but I am not sure if that is right

or is it the same?

Zacharý

before

z = (x - mu_x)/sigma_x

Lembik

ok so take v = [2,4]. We get [-1,1] and we divide both by 2^2 +4^2

that doesn't seem to have unit length

I mean divide both by sqrt(2^2 + 4^2)

Zacharý

4:21 PM

Sorry, I might have confused disciplines there >_<

Lembik

we -1/sqrt(20), 1/sqrt(20)

but that isn't of unit length is it?

Erik the Outgolfer

@ngn ...(e.) .NET object, (a.) array/niladic function/namespace/class

Zacharý

Subtract the mean, then divide by the range.

Lembik

any help much appreciated

@Zacharý so what do you get for the vector [2,4]?

oh we divide by sqrt(1 + 1) ?

Zacharý

[2,4] => [-1,1] => [-1.1]/(1--1) => [-1,1]/2 => [-.5, .5]

Lembik

4:25 PM

thanks

Zacharý

Range = Max - Min

Adám

Want to learn #APL? Join Adam tonight at 18:30 UTC for another Informal #APL learning session in The APL Orchard https://chat.stackexchange.com/rooms/52405/apl See https://chat.stackexchange.com/transcript/message/41299896 … … … if you don't have 20 Stack Exchange rep points

Tweeted by dyalogapl on March 14, 2018 at 3:47 PM

user202729

@ngn Is (n-1)%(l-1) th link (0-indexing) better?

@EriktheOutgolfer I believe what I wrote is correct. (exclude the main link)

DJMcMayhem

@Adám Is your job basically just advertising and teaching APL?

J. Sallé

@DJMcMayhem He's also a powerful wizard wielding the APL grimoire.

5

DJMcMayhem

4:33 PM

@Calvin'sHobbies @Downgoat imgur.com/gallery/XIvFv

user202729

I prefer J. One language with SBCS is enough.

(Jelly)

Adám

@DJMcMayhem No, I mainly develop and maintain internal and external tools. I also provide customer support, and help develop the language.

@user202729 There would be no Jelly if not for APL.

DJMcMayhem

Ah, OK. I assume you mean the parts of the language that are written in APL, rather than the C part?

Adám

@DJMcMayhem No, I mean coming up with ideas on how APL should be extended and made faster.

DJMcMayhem

So like new atoms?

Adám

4:37 PM

@DJMcMayhem Yes. E.g. I came up with ⊆ which was added last year.

@DJMcMayhem I also help my co-workers with APL. E.g. the CXO just now IM'ed me: "need help with APL :-(" "I'm feeling too stupid this afternoon"

DJMcMayhem

Haha

Dennis

@ConorO'Brien I have no idea what you were even talking about, but your chat message inspired me to create Dodos.

6

DJMcMayhem

Hello world

@Dennis I am truly terrified of this language

J. Sallé

@Adám sounds like me in The APL Orchard. Totally relatable.

Pavel

@Dennis wat

Erik the Outgolfer

4:44 PM

@DJMcMayhem primitives :P

Dennis

@DJMcMayhem There ought to be ways that are much golfier, but I wanted to implement decimal literals.

DJMcMayhem

@EriktheOutgolfer Are you calling me primitive?

:P

ngn

@user202729 I don't speak Jelly. I'm trying to understand what this means. Is it really like that: (n-1)%(l-1) ? So you can't call the last "link"?

Pavel

So what is this lang?

The readme isn't that clear

Dennis

Docs are under construction. I'll try to get the basics up this evening. The main idea is that f(n) can call f(n-1), but if it tries to call f(n+1) there's an exception and n is returned unaltered.

ngn

4:46 PM

@EriktheOutgolfer dyalog's oop extensions are not apl, syntactically they and the schizos are a can of worms

Dennis

There are three builtins: dab (remove first argument), dip (decrement the arguments if positive, increment if not), and dot (dot product with vector of 1's, aka sum).

Erik the Outgolfer

@ngn that depends on what flavor Zacharý is talking about, for example if it's specific to Dyalog then those should be a part of it, but, in general, those are not supported

user202729

@ngn Yes. The las5 link can't be called thid way. (there are some examples ti hekp understandingz)

Dennis

There is no control flow, there are no literals, Only function definitions. The function at the top (main) takes input and returns output.

Erik the Outgolfer

@Dennis why is that language based on D

(it's evident because, well, dot)

Dennis

4:49 PM

Because dodos.

user202729

@Dennis So it will always halt?

Zacharý

@EriktheOutgolfer Not worried about OOP.

@EriktheOutgolfer Huh?

Dennis

@user202729 Yes. Loops can be arbitrarily large (at least in theory, there's no TCO yet), but since the set of possible inputs is well-ordered, they will be finite.

Conor O'Brien

@Dennis I could have phrased the cmc better, but this language looks fantastic

Pavel

@Dennis What's TCO?

Conor O'Brien

4:52 PM

tail call optimization

New Sandboxed Posts

0

A: Sandbox for Proposed Challenges

code-golf Introduction Humans are a remarkable species, but we can be very awkward to understand sometimes—especially for computers. In particular, we seem to like writing polynomials in a very convoluted fashion with seemingly arbitrary rules. What is the shortest program you can write to f...

Dennis

I'm talking about Dodos, not Jelly.

user202729

@ngn Wrong reply target.

No, l is the total number of links in the whole program, not number of links before the current.

ngn

@user202729 I am confused...

@user202729 so, again, it's impossible to call the last link then?

Dennis

Yes, the main link cannot be called by these.

Ç, Ñ, et al. work though.

user202729

4:59 PM

But (1) it's the main link(TODO add this) and (2) as Dennis said above.

ngn

ok, I think I got it

so in the jelly language itself link numbering is 1-based

Dennis

All indexing in Jelly is 1-based.

ngn

@Dennis ugh...

@Dennis I thought it was based on J and "[]IO delenda est" long ago...

DJMcMayhem

Chat is freaking out on me

Dennis

Time travel?

DJMcMayhem

5:08 PM

I guess

NieDzejkob

Is your computer's clock set correctly?

DJMcMayhem

Does that mean the apocalypse happened? That's the only reason I can think of TNB being quite for 11 hours

@NieDzejkob Oooooh, no it isn't. My clock didn't correctly update for time zone (no idea why), so I set it manually, but for some reason it went to 11 PM instead of 11 AM

ngn

@DJMcMayhem someone has implemented "hours ago" as (x-1)%12+1 :)

DJMcMayhem

Sorry, s/time zone/daylight savings

J. Sallé

@DJMcMayhem for some reason, on my PC, Discord's clock is exactly 2 minutes ahead of mine. So when my clock is at 23:58, messages are shown as being sent "Tomorrow 00:00"

Conor O'Brien

5:11 PM

for challenges involving randomness, is it alright to take a seed as input? I couldn't find a meta concerning this, tho I might be just incapable of searching :P

DJMcMayhem

@ConorO'Brien This sorta came up with one of pavel's recent answers, but there's no official answer on it as far as I know

I think it would be worth a meta post. I'd be happy to write it up if you don't want to

Conor O'Brien

I can write it up

DJMcMayhem

Points to hit on: 1) Can PRNG be unseeded? 2) Can "random == user input"? 3) Can "seed == user input"? 4) Can "Random == time"?

Adám

@DJMcMayhem can you edit CTO to CXO? I feel bad to speak about the wrong person.

DJMcMayhem

Done. (I know it's done because I essentially just pinged myself)

@ConorO'Brien Example of #2

Adám

5:16 PM

@DJMcMayhem Thanks.

Conor O'Brien

@DJMcMayhem thanks, these are good points, I'll make sure to find examples of each

DJMcMayhem

@ConorO'Brien And Example of #1

I can't think of any for #3 or #4

Conor O'Brien

I know a few for #4

@DJMcMayhem I am slowly realizing that I may not have sufficient time to make a thoughtful post. I have some progress here, do you mind picking up on it/rewriting it?

DJMcMayhem

Sure. Are the first two links supposed to be the same?

Conor O'Brien

oh probably not no lol

ngn

5:23 PM

@Adám if a C-whatever-O says "i need help" or "can i show you something?" - run away, most likely your code isn't working :)

Conor O'Brien

@DJMcMayhem I think a persuasive point to make would be C's srand boilerplat and languages like BF which can implement PNRGs but not time

DJMcMayhem

OK, I'll make sure to mention those. I was already thinking about brain-flak for one

Adám

@ngn Nope, he needed help with how to use a new primitive. Also, it is usually me who goes to C?Os to "show something".

J. Sallé

@Adám ⍤ new primitive!

Adám

@J.Sallé New in 16.0: @

Zacharý

5:37 PM

My font renders rank ⍤ weird in non-monospace environments...

Adám

@Zacharý How does it look?

betseg

CMC: 426880√100005/13591409 (first term of the Chudnovsky algorithm for calculating pi)

Zacharý

○ yeah, happy pi day!

The ¨ is raised up a bit too high, it is the same high as if it were ¨.

*height

DJMcMayhem

@ConorO'Brien Any links to #3 or #4?

betseg

CMC2: for input i, calculate 426880√10005 / Σₙ₌₀ⁱ (6n)!(545140134n+13591409)/(3n)!(n!)³(-262537412640768000)ⁿ

Adám

5:41 PM

CMC: Given a matrix where the second column has only 1s and 2s, replace 1s with "dir" and 2s with "file". E.g. [["Prog",1,"Hello"],["Probs",2,"World"],["CG",1,"!"]] → [["Prog","dir","Hello"],["Probs","file","World"],["CG","dir","!"]]

dzaima

@betseg shouldn't there be one less 0 in the root?

J. Sallé

@Adám ah, and here I was getting all excited for new primitives D:

betseg

@dzaima yep. too late to edit now.

Adám

@Zacharý Screenshot?

@J.Sallé None in 17.0, sorry. Hopefully in 18.0, together with even more exciting stuff.

Zacharý

dzaima

5:43 PM

@betseg SOGL, 15 bytes

Zacharý

@Adám But TAO is a thing in 17.0, right (I know it's not a new primitive)?

Adám

@Zacharý Yes.

@Zacharý Not pretty, but at least it is very clear what it is. Can you screenshot (screenshoot?) ¨~⍨∘⍤○⍥?

Conor O'Brien

@DJMcMayhem 4: codegolf.stackexchange.com/a/158005/31957

Zacharý

Conor O'Brien

another 4: codegolf.stackexchange.com/questions/114520/flip-a-coin-for-me/…

Zacharý

5:48 PM

It's only in non monospace environments that ⍤ is weird

DJMcMayhem

I actually found an answer for #4, so this post won't need to address that

Adám

@Zacharý Can I see them in monospace?

Zacharý

The snapshot I showed you is the monospaced one, I think

Adám

@Zacharý But it doesn't have ¨ high over ∘ like your previous screenshot.

Zacharý

The LATTER snapshot is monospaced, while the former is not. It's weird in non-monospaced environments only.

Adám

5:53 PM

tfw you rep graph is a bottom-right 90° angle:

DJMcMayhem

0

Q: What are the standard requirements for answering a [random] challenge?

There are lots of of challenges involving randomness. However, not all languages have a built-in method for randomness. Some languages, like python or most golfing languages make it really easy. You can simply call some built in function that will return a pseudo-randomly generated number. There ...