The Nineteenth Byte

totallyhuman

7:00 PM

@Zacharý :P

Peter Taylor

@orlp, no. Do I not have a test case which gives a counterexample? If not, I'll add one.

orlp

you only have two cases with n = 1

Zacharý

I'm tempted to flag for some reason ... won't make a mistake like that again.

orlp

1 [2 4 10] 6 this is solved by tossing the minimum into every chute

Peter Taylor

Right, on it

orlp

7:01 PM

1 [3 5 5 5 5 5 5 5 5 5] 5 this one by tossing the maximum into a single one

Peter Taylor

1 [4 5 15] 10

orlp

alright

that one is similar to 'toss the minimum into every chute' but with some sort realization that throwing one extra allows you to rule out another chute

@PeterTaylor do you have a program that solves it? I have a feeling there is an elegant simple solution but I can't think of it at atm

Downgoat

@ConorO'Brien I think it might be better to split functional operator and operator functions into two seperate proposal so to increase the chance at least one will get in :P

Peter Taylor

I have reference solutions in Java, Python, and CJam, although only the CJam one is golfed

orlp

@PeterTaylor are they big (complex) or fairly simple ones?

Zacharý

7:08 PM

@PeterTaylor Well ... yeah, it's Cjam, a golf-lang.

Peter Taylor

There is an elegant solution, but it's tricky to prove it correct.

orlp

I think there's a condition you can use for a recursive solution in that there is no observable effect of throwing less than the minimal number of balls, so you can throw that number into every chute and recurse on the resulting configuration

that sounds like a combinatoric explosion though, unless there's some dynamic programming substructure

Peter Taylor

Major hint: there's a dynamic programming solution in quadratic time.

Zacharý

Hey, I'm wondering if I should start developing my codepage right now.

Downgoat

@ConorO'Brien wait if JS don't have operator overload is functional opeartor really useful

totallyhuman

7:15 PM

also there needs to be more groovy on ppcg

groovy is pretty... groovy

orlp

@PeterTaylor I just had a really interesting mind experiment reasoning flaw

Zacharý

@totallyhuman Dang it, that's MY thing. (because MY is my programming language)

orlp

'surely the worst case is always throwing the ball in the biggest chute on the first throw'

Downgoat

@ConorO'Brien Question: what would happen if you did (+)()

orlp

'oh - then the algorithm is simple, throw 1 ball in your first choice then always switch after that'

New Sandboxed Posts

7:20 PM

0

A: Sandbox for Proposed Challenges

Play a game of Flow Flow Free is a game where you have to join different coloured points on a board with a "pipe", in such a way that no pipes cross and that 100% of the board is covered in pipes. Your task is to take an empty Flow board and output the solution to it. Each board is guaranteed on...

Peter Taylor

Allow me to correct myself: it's cubic, not quadratic, in the worst case.

Zacharý

Anyone else hate O notation?

orlp

@Zacharý why?

Conor O'Brien

@Downgoat probably an error

@Downgoat not entirely, but we could anticipate overload in the future

Zacharý

I just ... do, it just seems so confusing (it's probably not, to be honest)

Business Cat

7:30 PM

I hated O notation when they made me work with the mathematical definition in school. Now it's just "oh, that's clearly O(whatever)"

Conor O'Brien

@Downgoat agreed

Business Cat

The mathematical definition is evil

Zacharý

@BusinessCat Yeah, I am a math-oriented person I was going into it thinking ... oh, this should be fine... later: NO NO NO NO NO NO NO NO

orlp

@PeterTaylor by the way you might want to implement some anti-bruteforce clause into your challenge or else I'm sure someone will iterate over every permutation of chutes and every possible ballthrowing strategy

Business Cat

Big O is much nicer when you can look at something and be like "oh, the biggest term is 5x^2 so it's O(n^2)"

Zacharý

7:34 PM

(I really wish Comp Sci AB was still a thing, I would've learned something)

Business Cat

AB?

Zacharý

AP Computer Science AB.

Peter Taylor

@orlp I think that's actually more complicated than some better approaches

And I think that most golfed approaches will be non-polynomial, which makes it hard to write an anti-brute-force clause

I do normally favour them, but I deliberately didn't include one this time

orlp

I'm trying to reduce the combinatoric explosion at the moment but I'm having a hard time figuring out what I can assume about the worst case

see also my funny flawed thought train above

Zacharý

strawpoll.me/13402345

Peter Taylor

7:40 PM

@Zacharý Category error

Zacharý

What?

orlp

@Zacharý What is your favourite, a banana or solar panels?

6

Zacharý

>.<

Lembik

people keep on downvoting codegolf.stackexchange.com/questions/130337/… but no one says why

hi all

Wheat Wizard

The time limit seems makes this a chameleon challenge - it becomes more about writing fast code than about writing short code. — Mego Jul 7 at 12:58

That would be my guess

Lembik

7:44 PM

@WheatWizard ok. that comments was a few days ago and the downvotes are current

Wheat Wizard

Ok then I don't know.

totallyhuman

0

A: Fibtraction - Fibonacci but with subtraction

Groovy, 27 bytes f={it<3?it:f(it-2)-f(it-1)} Try it online! Okay, so I feel like I stumbled upon a great language for golfing. This code is basically what most other answers are doing, if n is lesser than 3, then return n or else return f(n - 2) - f(n - 1). But there are a few cool things abo...

Lembik

@WheatWizard it's tricky as I asked about this on codegolf.meta.stackexchange.com/questions/13231/…

and the answer looks clear to me that it's ok how it is

it's almost like people have their own minds :)

maybe I should link to that from the question??

Wheat Wizard

My guess new is that people read it, thought it would be easy, did it assuming the time limit would be a non-issue and then realized how hard it was to make a valid answer given the constraint. I feel that the issue of it being a chameleon challenge is mostly that the time constraint is the most important aspect of the challenge but is not stressed as so in the body of your question.

Lembik

@WheatWizard interesting.. so a change of emphasis might help

Wheat Wizard

7:50 PM

I think so

Lembik

added a new sentence near the beginning

also added a link to the meta question

Wheat Wizard

I would like to try the question but I don't have a "modern desktop" so I would not be able to reasonably test my solution.

Lembik

@WheatWizard I don't think that's right

if you use a suitable algorithm your code will be almost instant

which will be the same on a modern desktop

Wheat Wizard

But its codegolf, so a suitable algorithm is not always the best one

Lembik

try orlp's code? It takes < 3ms

@WheatWizard I am not sure there exist algorithms of intermediate speed for this problem

I think you are either very fast or very slow

I did think about this before posing the question :)

I do fear for the prospects of my next challenge which will be the discrete logarithm

ok there is an intermediate speed solution which is to use discrete logarithm... which is bonkers :)

orlp

8:04 PM

@PeterTaylor I have a solution I believe to be correct but that's too slow for the bigger testcases

gist.github.com/orlp/f40597ca35d0bbf48adb90e6d0dc4a67

Peter Taylor

It has one parameter but the recursive call passes two...

orlp

@PeterTaylor reload gist, I edited something

(I moved goal out of the recursive call but forgot to remove it there)

actually, after implementing caching now (after noticing I can sort the arrays) I find that it's incorrect on bigger cases

Christopher

0

A: Sandbox for Proposed Challenges

Make me some ASCII cities! Your job is to pseudo randomly generate ASCII cities (top down view) with n building per row with x rows. Each house should have a number label starting with 000 and working up. First number is the row (assume x<9) and the other two are places in the row. i.e 000, 001,...

How fast would you VTC that?

tags: code-golf, ascii

totallyhuman

orlp

@PeterTaylor actually that was a minor hiccup, I think I solved it now

Peter Taylor

8:18 PM

The overall structure looks plausible

orlp

now I need to golf it

DJMcMayhem

2733

Q: Proper use cases for Android UserManager.isUserAGoat()?

I was looking at the new APIs introduced in Android 4.2. While looking at the UserManager class I came across the following method: public boolean isUserAGoat() Used to determine whether the user making this call is subject to teleportations. Returns whether the user making this call ...

Zacharý

Unicode ... why???
℘ = script capital p

totallyhuman

what's wrong with it

@DJMcMayhem yeah but now we can confirm if downgoat is sheep in disguise

orlp

@PeterTaylor may we take the pipes in descending order?

Peter Taylor

8:30 PM

> You may also assume that the counts are sorted (ascending or descending), as long as you state that clearly in your answer.

orlp

oops :)

Zacharý

𝒫 = MATHEMATICAL SCRIPT CAPITAL P

Adám

@Zacharý There are lots of errors in Unicode.

Zacharý

Still... how do they mistake a lowercase for an uppercase?!

Adám

@Zacharý UP TACK: ⊥ APL FUNCTIONAL SYMBOL DOWN TACK UNDERBAR: ⍊

@Zacharý How do they mistake up and down‽

Zacharý

8:34 PM

Wow. Just wow. They hate APL. alot.

Peter Taylor

@Zacharý, I'm not sure why you're blaming the Unicode consortium for the implementation of your font provider

Adám

@PeterTaylor No, Unicode have admitted it was a mistake.

totallyhuman

welp my meta post failed ;-;

Zacharý

And that they won't fix it ... (I get the case of the ℘, but SERIOUSLY WHY NOT CHANGE ⍊?)

Peter Taylor

It looks pretty upper case in most fonts here: fileformat.info/info/unicode/char/1d4ab/fontsupport.htm

Zacharý

8:39 PM

There fixed...

Adám

@Zacharý All the APL tacks are like that. See footnote *** here

@Zacharý On the other hand, there was once a dentist who liked APL a lot. I've not been able to find any info on these glyphs' origin or use.

Zacharý

Some dentistry symbols there make sense (the corner ones).

Actually, it's probably coincidence to be honest (very generic symbols)

Adám

@Zacharý Sure, and some are basic geometric shapes. However, some seem eerily, specific, no?

Zacharý

Where does iota even come from to be honest?

Christopher

Lets see if my multi month long sandbox with 1 then 0 score gets closed. (pop con fyi)

orlp

8:45 PM

@PeterTaylor I posted my answer

Adám

@Zacharý index, indices, in

Oh. Reverse?

@Zacharý huh?

⌽

@Zacharý mirror over the axis indicated, just like ⍉ and ⊖

Zacharý

8:48 PM

Okay, any explanation as to why unicode considers APL alpha,omega,iota different from the greek versions? Just 'cause?>

Uriel

@Zacharý Probably bacause they are used as APL function and dfn arguments and not greek letters

Adám

@Zacharý Because they serve different roles from the word characters.

Uriel

you'd want to have a clear distinction for these

Zacharý

Okay, I think this is the last one: why the difference between the small membership (GNU IIRC) and normal membership (Dyalog IIRC)

Uriel

I thinks there's also delta

Zacharý

8:51 PM

Delta could just be a generic triangle to be honest,

Adám

@Zacharý Not the APL symbol. It is a word character.

Peter Taylor

@orlp, does functools work for you? I get an error: unhashable type: list

orlp

@PeterTaylor sorry, my usage section was incorrect

it expects a tuple as input

e.g. S(2, (2, 4, 10))

DJMcMayhem

CMC: Write a full program, that when ran will print the current number of moderators on the SE network. (545 right now)

Zacharý

How does this codepage look? bitbucket.org/zacharyjtaylor/my-language/src/…

Uriel

8:56 PM

@Zacharý too much APL. too much unbalanced width unicode chars.

Adám

@DJMcMayhem 545

@Uriel too much APL ‽ Traitor!

DJMcMayhem

No, it should print the current number at the time of running it. So running it once more mods are elected it should still show the correct number

Zacharý

Only a few are APL actually.

totallyhuman

anybody know pyth

never mind o worked it out

Zacharý

Most are just inspired by APL. The quad and quote quad is input only. The squish quad is just empty array.

Adám

8:59 PM

:38694150 I think you got it right this time.

@Zacharý Why not ⍬?

Zacharý

Thanks ... forgot about that one!

New Main Posts

-1

Q: Make me some ASCII cities!

Your job is to pseudo randomly generate ASCII cities (top down view) with n building per row with x rows. Each house should have a number label starting with 000 and working up. First number is the row (assume x<9) and the other two are places in the row. i.e 000, 001, 002... Every house should v...

ascii-art code-challenge generation

Zacharý

√ and ¯ are two separate chars.

orlp

@PeterTaylor I added a small explanation of my algorithm

Adám

@Zacharý Let me guess: nth root and negative?

Uriel

9:01 PM

@Zacharý for reference, my (partial) codepage: ⍝æЂĳξøœßþÿπ¶±¿¤≉√«»⣿⌺♜♒⽂※‼᧻ಠ🦄🤢🛰️⚽🎷𐩓׆

Zacharý

No, square root and conjugate. (_ is negative)

Adám

@Zacharý Uh, J. What is monadic +?

@Uriel Unicorn?

Uriel

@Adám the imaginary unit.

totallyhuman

wait anybody know pyth?

Uriel

@totallyhuman yup

Zacharý

9:03 PM

Stack-based. + is only dyadic (will be when I implement it)

totallyhuman

@Uriel FN[2,3,12,15,22,25,32)yN

Adám

@Uriel You choose the glyphs for the entertainment value, right?

Zacharý

You should have U+1F595 exit the stack <troll face>

Uriel

@totallyhuman what are you trying to achieve?

totallyhuman

shouldn't that apply the function y to every n in the list?

orlp

9:06 PM

@PeterTaylor is that the same algorithm you came up with?

Adám

@Zacharý Why don't you have a Cap like J? It will push a special null value causing the next dyad to act as a monad, thus giving you a host of available monadic functions, each (possibly) related to its dyadic cousin?

Peter Taylor

It's similar to one of my algorithms, specifically the first one I proved correct.

Uriel

@totallyhuman no, I think that would be My[...] (M for map)

Adám

@Uriel Nun Hafukha. Nice.

Peter Taylor

I actually conceived it as a minimax tree search, and you skip the max nodes by effectively arguing that they're unnecessary

Zacharý

9:08 PM

If I were to make a language that complex, I would spend it on something other than a golf-lang (that's reserved for Jelly)

totallyhuman

ohh nvm pyth doesn't use commas in lists

Peter Taylor

I've just tried golfing my Python solution and got 152 bytes with no need for lrucache, but it took me several hours of refining my ideas to get to that algorithm

Uriel

@Adám 🛰️ is currently supposed to reflect part of the code on another part of it, when the code is wrapped around an elliptic surface, based on some solar and lunar calculations.

Adám

@Zacharý How does it make the language any more complex? Just push 234154236526432.5431652753865375e531 and use that as special value.

Christopher

oeis.org/A257187

Zacharý

9:11 PM

What if someone WANTS 234154236526432.5431652753865375e531?

It's my first attempt full at a language, so I'm keeping it like it is, pretty much no syntax/special stuff.

Adám

@Zacharý Can't you just use null or NaN?

@Zacharý Fair enough. It is YOUR language after all.

orlp

@PeterTaylor my solution isn't particularly golfed, although the solution isn't particularly nice to golf in Python either

New Main Posts

0

Q: Create unique ids

You are to write a program that takes a list of strings as input. For every string in the list you are to determine the smallest N such that no other string in the list begins with the same N characters as the string in question. Now compose a new list with all these strings. Here is an exampl...

code-golf string

Adám

@Zacharý Let me know if you want glyph inspiration.

Uriel

@Adám

⌶ ⌷ ⌸ ⌹ ⌺ ⌻ ⌼ ⌽ ⌾ ⌿ ⍀ ⍁ ⍂ ⍃ ⍄ ⍅ ⍆ ⍇ ⍈ ⍉ ⍊ ⍋ ⍌ ⍍ ⍎ ⍏ ⍐ ⍑ ⍒ ⍓ ⍔ ⍕ ⍖ ⍗ ⍘ ⍙ ⍚ ⍛ ⍜ ⍝ ⍞ ⍟ ⍠ ⍡ ⍢ ⍣ ⍤ ⍥ ⍦ ⍧ ⍨ ⍩ ⍪ ⍫ ⍬ ⍭ ⍮ ⍯ ⍰

?

Adám

9:16 PM

@Uriel Yes sir! Actually, what?

Zacharý

Basically none of those are used in APL ... that's the sad part.

APL portion of Miscellaneous Technical Unicode block.

Adám

@Zacharý Why is it sad. MY doesn't have to look like APL. What do you e.g. plan for sort up and down?

@Zacharý There is actually a way to ask APL which symbols it knows about (even secret experimental things!), but I'm not allowed to tell you.

orlp

@PeterTaylor regardless of how golfable or optimal my algorithm is, I'm pretty happy I did solve it :)

Zacharý

⍋⍏⍒⍖ are for sorting.

orlp

is this a well-known problem?

Zacharý

9:19 PM

(Time to brute force everything in APL ... wow)

Adám

@Zacharý ⍒⍋ are for grading. Why not ∇∆ for sorting?

Uriel

Zacharý

∆ for increments.

Adám

@Zacharý Ah. I've proposed <> for sorting. Alternatively ∨∧.

Uriel

@Adám why? you can just index after ⍋

Zacharý

9:21 PM

Because code golf.

Adám

@Uriel materialize?

Zacharý

One use of squish-quad (I think that's OO or something...)

Adám

@Zacharý Shhh!

@Zacharý Right, but what does Uriel mean? Indexing (dyadic ⌷)?

Mr. Xcoder

@orlp I think your solution is 201 bytes, not 198

Zacharý

@Adám What? Do they NOT know you probably will use it for golf?

And ⌷ doesn't work with array indexes, does it?

Adám

9:23 PM

@Zacharý They do.

@Zacharý It does. But if you want to sort anything you have to be clever.

Zacharý

They're even having a dang workshop for it at Dyalog '17.

Uriel

@Adám sorry, meant like k[⍋k]

orlp

@Mr.Xcoder you forgot the tab <> spaces trick in Python

those 4 spaces are 1 tab

Adám

@Zacharý Yes, and I didn't even have anything to do with that suggestion. Honest!

Zacharý

But ... of course you're going to be in it.

Adám

9:25 PM

@Uriel Won't work neither on matrices, nor on vectors of vectors.

@Zacharý Yes. Orders. I had no influence on that either.

Uriel

@Adám how do you even define sorting matrices?

Adám

@Uriel Lexicographically: {⍵[⍋⍵;]}

Zacharý

APL-style. Rows. Columns. etc.

@Adám That definitely was indirect influence, considering ... PPCG

Adám

@Zacharý Stefano and Nick were here long before me.

Uriel

@Adám and morten did some SO

Adám

9:30 PM

@Uriel Oh, on SO there are a bunch of APLers.

Zacharý

I was referring to you participating in the workshop.

Adám

@Zacharý Oh, ok. Yeah. Wonder what we'll do…

Zacharý

Wait ... did Stefano work for Dyalog?

Peter Taylor

@orlp I don't think so. As mentioned in the sandbox post, it's an anonymised version of a question which seems to be from an Amazon interview a couple of years ago, but I could find barely anything using the phrasing of the Amazon version

Uriel

@Adám and most of their answers looks like dyalog promotional

Zacharý

9:32 PM

v16 stuff. (I've been to the page for '17 workshops)

Adám

@Zacharý Yes, long ago. He's with APL (Italiana) now.

Zacharý

No wonder he beat Dan B. by one byte in the year game last year.

Adám

@Zacharý Oh. I didn't look. Interesting though. Dan brute forced.

I don't know if Stefano also did, but I know that quite a few of us did.

Zacharý

I have a feeling Jonas Stensiö and Veli-Matti Jantunen brute-forced as well (tied for first).

Adám

@Zacharý They cheated. Finns!

Zacharý

9:37 PM

Is that like an inside joke or something?
Who would want to cheat, and rob Dan of his posthumous win?

Adám

@Zacharý No. They didn't know how Dan would fare. Also, they did cheat. They got results which looked correct, but were character vectors instead of numeric arrays. That was against the spec from the beginning.

Zacharý

(I was referring to the "Finns!")

Mr. Xcoder

@orlp I figured that out in the meanwhile

But I am too tired

It's 12:40 AM here

Bye

Adám

@Zacharý Oh, that's just a joke on both being Finnish. Not inside anything. Possibly Nordic humor.

Uriel

@Adám so are they Finns or Finnish? or is Finnish an adjective for Finns?

Adám

9:43 PM

@Uriel They have Finnish nationality, so they are Finns. Just like I'm a Dane because I'm Danish. And Swedes are Swedish citizens.

Zacharý

@Adám Denmark. Canada. UK. Anywhere else?

Uriel

@Adám but Dan was also Dan-ish and he wasn't a Dane.

Zacharý

@Uriel Does this have to do with APL or what? (the puns)

Adám

@Uriel He wasn't Dan-ish, he was just Dan. Daniel, actually. His family always called him Daniel. He did live in Denmark though, and learned some Danish.

Uriel

@Zacharý no, he was Danish as you are Zacharish

Zacharý

9:45 PM

@Uriel I even noted "(the puns)"

Adám

@Uriel But I am an Adam (aval ani adam echad)

His wife and daughter visited me in London a couple of days ago. And today we looked at an old bug report that Dan had filed, and needed to ask what exactly he meant. Miss him. ⊥⍨

Uriel

@Adám May he rest in peace.

Zacharý

^

Mr. Xcoder

@Zacharý Name change?

Uriel

@Adám btw, wasn't his memorial challenge the first one?

Zacharý

9:53 PM

@Mr.Xcoder ‽ (had that on clipboard)

Adám

@Zacharý You changed username.

Zacharý

Oh. Yeah, I changed my username.

Adám

@Uriel the first what?

Uriel

first memorial challenge

Zacharý

wut. repl.it/JVhT/1

DJMcMayhem

9:54 PM

I'm super excited about this sandboxed post. Does anyone have thoughts on it?

Adám

@Uriel Are there more?

Zacharý

@Uriel Are you referring to the first memorial being the challenge?

Uriel

@Adám Chuck Berry, Adam West

Adám

@Uriel Oh. Possibly.

Zacharý

Oh, the first memorial challenge being posted on PPCG.

Grrr. English ambiguity.

Adám

9:57 PM

@Zacharý Sure, a dyadic function that doesn't use its arguments, but just returns a constant value. What's wrong?

New Sandboxed Posts

1

A: Sandbox for Proposed Challenges

Make your language mostly unusable cops-and-robbers arithmetic Inspired by this comment... In this challenge, you are tasked with running some code that makes it so that your language no longer satisfies our criteria of being a programming language. In that challenge, that means making it so...

Zacharý

The fact that no spaces are needed between the interrobang and the arguments.

Uriel

@DJMcMayhem looks good, but this can have really complicated workarounds for some languages

Zacharý

I would expect it to think ‽2 was a variable.

DJMcMayhem

@Uriel So... That's good for cops? Or good for robbers? I'm slightly worried that it'll be too hard to get a safe cops answer.

Transcript for