The Nineteenth Byte

Alex A.

1:00 AM

@PhiNotPi Nothing wrong with that.

PhiNotPi

Like, I literally wrote this today: soundcloud.com/phinotpi/serene

Still WIP, but it's more music than the average person wrote today.

Sp3000

The average person writes 4'33", every day :)

4

Alex A.

Just when I was hoping someone would make a John Cage reference.

@PhiNotPi Not sure what it is about it (maybe pentatonics?) but it sounds vaguely Japanese to me. I really like it, nice work!

PhiNotPi

Here's a death metal cover of 4'33: youtube.com/watch?v=hUzI3Ui1Eok

Alex A.

1:18 AM

He nailed it.

PhiNotPi

I also expanded one of the other pieces I was working on. soundcloud.com/phinotpi/cave1b

Alex A.

That's fantastic!

PhiNotPi

thanks!

Alex A.

You've got some serious talent there.

PhiNotPi

1:40 AM

well, I keep working on it

Pietu1998

5:12 AM

Just a random question that came into my mind: what's the shortest CJam quine not involving leaving code blocks on stack? AFAIK the {"_~"}_~ one relies on printing the code block as is, followed by the string.

Sp3000

There's the example one in the online interpreter - basically the same thing, but it isn't a block

Pietu1998

Right. Does there also exist one that does not use eval?

Sp3000

There might be one of the form <number> <number>b:c if you can be bothered checking if a solution exists

Pietu1998

So that'd basically base-convert a number to itself?

Sorry, I'm not very experienced in CJam although I'd like to try more sometime

Sp3000

5:33 AM

Hmm too bad CJam doesn't have a raw string format or copying the Ruby/Python quine might have been possible

Optimizer

6:17 AM

@Sp3000 I don't think so I wrote 4'33" yesterday at all

or day before

VisualMelon

I don't think we'd quite consider you average though, @Optimizer ;)

Martin Büttner

@Sp3000 do you know how the new e% works?

Sp3000

I think I do, but I couldn't find an option like %r in Python which would put quotes around strings

Martin Büttner

when I try to use it I get "Format string requires '%'."

Sp3000

What are you trying?

Martin Büttner

6:30 AM

was trying %s and %p and %%... apparently none of those exist... %d works

Sp3000

If you're formatting a single string like "String: %s" "abcd" e% that won't work because strings are arrays of chars, so you need "String: %s" ["abcd"] e%

Martin Büttner

anyway, here is an 8 byte no-eval quine: "_`o"_`o

@Pietu1998 ^

oh, apparently I've got exactly 500 answers now

Sp3000

Nice :) How many from answer chaining/CnR? :P

Martin Büttner

I don't think many are from answer chaining

CnR though...

48 from CnR

Sp3000

:P

Well I was thinking OEIS, which has kind of died down...

There's a lot of interesting sequences I'd like to do, but unfortunately there's too much noise as well

Martin Büttner

6:38 AM

@Sp3000 I still mean to complete that Fission program, but I don't think I'll post it to EOEIS. I think the sequence I picked would still be an interesting golf in itself, so I'll probably post it as an answer there when I get around to writing that challenge. (it wouldn't have been very deep in the tree anyway)

Martin Büttner

6:53 AM

@Sp3000 @Pietu1998 7 bytes:

"_p"
_p

Sp3000

Hmm that one looks familiar - did you post that before some time ago?

Martin Büttner

yeah I think I was

Sp3000

Yep

Glad to know Martin is consistent in thought patterns, even after 2 months

Martin Büttner

:D

in fact, I did think of the \ version first, but discarded it in favour of o right away this time :D

Optimizer

10:59 AM

anyone up for 2 byte golfing for the bean machine ?

PhiNotPi

I'm up... But I have to do something else for the next half-hour... :(

sirpercival

hi all

Optimizer

nvm, just did it

sirpercival

:( my tic-tac-toe thing is downvoted

Optimizer

I guess, its time for "Take that Pyth!" ?

sirpercival

11:10 AM

anyway, i expect you are all writing bots for aBOTcalypse? :D :D :D

Optimizer

12:00 PM

@randomra wow, that looks so like CJam

randomra

@Optimizer to me it looks like J or random signs :)

it does miss some J signature chars like .:@

Optimizer

@randomra Atleast it runs perfectly fine in CJam

given that CJam is not exception safe, that's something!

try

Sp3000

If you have ( or ) in your code try replacing by their hex values %28 and %29

Optimizer

@Sp3000 somehow it still worked

Sp3000

Usually I just replace anyway :/

Optimizer

12:05 PM

fear of the brackets

randomra

wow, what does it do? it seems nontrivial

lets see

Gibberish XD

yeah

I mean, :)~ who does that

randomra

12:10 PM

doesn't this make sense? q~]:)~ or is there a shorter variant?

Sp3000

Well it does things, just not something that'll come up a lot

btw speaking of CJam, cjam-docs.readthedocs has become cjam.readthedocs.org

randomra

sorry that my J code isn't an often used CJam code, will try harder next time :)

also, long-face is not implemented... RuntimeException: Long :) not implemented

Optimizer

:X works only on arrays

but nice idea to do an implicit , in there!

randomra

5 chars shorter than Jakube's pyth, that can't be right

Sp3000

Different languages excel at different questions :P

Optimizer

12:17 PM

If you have beaten Jakube, only Dennis can beat you now

randomra

@Sp3000 I don't think J has any advantage here and my code seems cumbersome, I'm sure Jakube's code will be halved eventually :)

Pietu1998

I'm kinda surprised the solutions are this long anyway, I've had the idea for my question for a while and to this point I've thought it'd be like 15 in some of the Great Golf Languages™

Not that I'd have a shorter one for now

randomra

@Pietu1998 I think too it should be around 15 chars based on my solution

Abdou Abdou

12:47 PM

salut :)

why did the op unlink all this post from my definitions ? codegolf.stackexchange.com/questions/50022/…

like sudoko-friendly , and pseudo-solvable , ...

trichoplax

I'm not sure what you are asking

Geobits

Probably because it reads better this way (IMO). The links went to Puzzling posts, not succinct definitions. Having the definitions clearly stated inline is less burden on the reader.

Sp3000

1:04 PM

I think they probably removed it because the definitions weren't really necessary for defining the challenge. Also in English "sudoku" is normally used over "sudoko" or "sudokoo", just as another note

Martin Büttner

who was the guy who knew enough TI-BASIC to double check Timtech's answers?

Geobits

Runer IIRC

Martin Büttner

ah yeah you're right

@Runer112 would you mind having a look whether this looks legit? (I don't see anything fishy, but I've grown a bit wary :D)

Geobits

Weary, wary, or both? ;)

Martin Büttner

:/

(thanks)

Runer112

1:30 PM

sure

Zgarb

Hello all

Runer112

looks like line 4 has a slight error

Martin Büttner

A wild Zgarb appeared.

Runer112

it should have 1+ before the second arg, like line 6

because indices in TI-BASIC are 1-indexed

Zgarb

I'm gone for two weeks and questions/day jumps from 4 to 5.8 o_O

Runer112

1:34 PM

strange that he got it right for one but not the other

Martin Büttner

@Zgarb which is surprising, seeing your question volume ;)

randomra

0

A: Sandbox for Proposed Challenges

Connecting Gaps with Tetris Pieces code-golf tetris You should write a program or function which given a list of tetris blocks as input outputs or returns the biggest gap between two points in the same height level which the pieces can connect. The 7 types of tetris pieces are the following: ...

will post if no comments

@Zgarb what timeframe is that calculated on?

I have posted like 10 questions in the last two weeks :)

Geobits

I think random got tired of CH getting all the question glory :P

Zgarb

@randomra Exactly 2 weeks, it seems

Sp3000

"and sleeps a bit between each line of output you get a -10 bonus." <-- :/

Zgarb

1:38 PM

I'll just post a quick golf and get back to work, things are crazy right now

Martin Büttner

@Zgarb definitely prioritise posting that challenge then.

Zgarb

@MartinBüttner Posting a challenge counts as a coffee break :P

randomra

did the answer ratio went down by the question volume increase?

Geobits

Not noticeably, it's still > 9:1

Optimizer

its so nice that even though most of Zgarb's challenges involve graphs and grids and structures and what not, but the input is simply an array of instructions :D

Geobits

2:22 PM

I have the feeling that this is not the intended answer. But screw that, it's a bear.

Martin Büttner

it won't

why?

sirpercival

ugh, how do i translate a list comprehension to pyth

Optimizer

asdkjhksad@#$asd!@sad

not the shortest way.

Martin Büttner

hm, there must be a better way to turn an array into a list of its non-zero indices

TheNumberOne

@Geobits How to have a spoiler. I know the answer.

Geobits

2:35 PM

>! spoiled text

sirpercival

i might know the answer as well

Geobits

I gave the answer. If you don't want a bear to be shot, don't surprise a hunter by placing one in his hut :P

trichoplax

I think the hut is pretty big. If you can leave it, walk 1 kilometre North, then 1 mile South, and still be in the hut.

sirpercival

@trichoplax nah it determines where the hut is

trichoplax

I'm guessing over a third of a mile North to South

@sirpercival It would if it was 1 mile North then 1 mile South

sirpercival

2:41 PM

...no?

trichoplax

But a kilometre is less than two thirds of a mile

TheNumberOne

My internet stops right after I post the answer.

sirpercival

ha i missed that

trichoplax

:)

TheNumberOne

@Geobits Does my answer make sense?

Geobits

2:53 PM

I thought about something similar, but using a slightly different definition of "bear" (see urban dictionary).

Rainbolt

Gross

TheNumberOne

It is impossible to increase your latitude by 1 km, and later reduce it by 1 mile, and end up at the same place.

Rainbolt

@Geobits I have two friends that attend the Texas Bear Round Up each year

Optimizer

@MartinBüttner better than ,,:) ?

Geobits

@Rainbolt Google images gives very interesting results for that.

Optimizer

3:05 PM

what about ,{)}/ ?

Rainbolt

@Geobits I don't even want to know

Geobits

I didn't scroll through much, but what I saw was generally SFW, for what that's worth.

Optimizer

@Geobits maybe he has content filter off ?

Geobits

Nah, I always have the content filter off, so that shouldn't make much difference.

Martin Büttner

@Optimizer I don't think we're talking about the same thing

[0 0 -2 0 1 123] --> [2 4 5]

Optimizer

3:09 PM

oh, that thing you are doing there at the end

little better:

:z:g_,,.*0-

but pretty sure that you need to change your algorithm to be anywhere near 20 bytes

even better:

_,,.e&0-

@MartinBüttner ^

randomra

how does the CJam , filter work? does it expect a function somehow? or two arguments?

Sp3000

ee{~{;}|}% awww {;} is too long :(

randomra

(I'm starting to confuse ><> and CJam operators... :/)

Optimizer

@Sp3000 mine is better ;)

Sp3000

@randomra Here's an example for getting all numbers which are equal to 3 mod 5: [0 5 2 2 5 7 13 3 8] {5%3=}, p

@Optimizer I never said mine was short :P

Pietu1998

3:21 PM

@Sp3000 Wow, a somewhat comprehensible CJam snippet

randomra

so blocks {} are pushed to stack as functions? without executing them

Optimizer

Martin is already working on a different algorithm..

ticket #64 is becoming interesting

Sp3000

Basically, although it differs slightly from eval in that it still need to be syntactically valid

randomra

and how cjam generally decides how many arguments an operator takes? based on the typeof the topmost stack element?

eg if its block, , is filter, not length/range

Optimizer

its decided by aditsu

randomra

3:25 PM

@Optimizer that's not helpful

Optimizer

I know right!

how can we let him decide everything

anyways, there is very little documentation of everything cjam. but the sourcefourge page should be a good start

Sp3000

I'm working on it :P Maybe I should do the operators instead of the tutorial first

Optimizer

@Sp3000 yes

Sp3000

As for number of arguments, I think they're the same for any single operator - its functionality just changes depending on the types

Optimizer

no operator has overloading in terms of number of operands

(except maybe ] and similar ones)

randomra

3:29 PM

I see, so the type of the first argument(s) decide the operator AND it's arity

Optimizer

the operator decides that

not the arguments

just special case blocks

{}X

randomra

but [1 2]{0=}, filters [1 2], gives length

so {} bonds to the operator?

Optimizer

34 secs ago, by Optimizer

just special case blocks

@randomra if it has one , yes

are you learning cjam to write a better J code that is more often used in CJam ?

randomra

@Optimizer but with your model this shouldn't work [0 1 5 0 6]{0>}0;, yet it gives 1 5 6

Sp3000

Guess I was wrong then :P

Optimizer

3:33 PM

@randomra everything is decided at runtime

(except f , : and . which are wired in at the time of syntax check)

Sp3000

Number of operands might be determined by types of topmost elements in that case

sirpercival

whee sandbox ideas

Optimizer

I still would say that treat block as special case. Would ease the understanding overall

randomra

@Sp3000 based on what Optimizer said the only deciding of arity is by the fact taht if topmost element is a block

Optimizer

yes, arity changes only for block (as far as I know)

I have no idea how I end up doing such typos

New Sandboxed Posts

3:37 PM

0

A: Sandbox for Proposed Challenges

Scralphabet A normal bag of Scrabble tiles contains the following letters (? is a blank tile, which can stand for any other letter): AAAAAAAAABBCCDDDDEEEEEEEEEEEEFFGGGHHIIIIIIIIIJKLLLLMMNNNNNNOOOOOOOOPPQRRRRRRSSSSTTTTTTUUUUVVWWXYYZ?? The letters have the following value: {"A": 1,"B": 3,"C": ...

randomra

it could be weirder like: arity is 1 if the second topmost element isn't an array :P

Sp3000

@sirpercival Oh god SESQUIOXIDIZING

You might want to provide a dictionary though

sirpercival

@Sp3000 just finished editing to add that in, actually

Sp3000

Awesome

sirpercival

do you like this challenge?

Geobits

3:44 PM

I like it, but you really should define/provide a single dictionary.

Sp3000

Single dictionary would be better, IMO too

sirpercival

ok i'll fix

done

Sp3000

As for like... I wouldn't be surprised if someone hits optimal quickly with a good heuristic, but I dunno

randomra

not short but did it _,,]z{~;g},{~\;}%

Sp3000

btw any special rules for blanks left over or do they deduct 0?

sirpercival

3:47 PM

they deduct 0

well, no

Optimizer

@randomra did what?

sirpercival

they contribute nothing to the letter values of the anti-word, but they increase its length

randomra

the nonzero positions

Optimizer

:D

randomra

time to download cjam so I dont have to use the mouse :)

the shell just quits on error? :O

Optimizer

3:52 PM

you saying something about mouse ..

also, on website, Ctrl Enter from anywhere runs

Martin Büttner

@aditsu ArrayList ArrayList e* not implemented :(

Optimizer

repeat corresponding times ?

Martin Büttner

@Optimizer ah that's neat, thanks

@Optimizer yes. so, like \:a.*, basically

actually

your code doesn't work

because I can't get the 0 index with that

Optimizer

:D

too bad

use Sp3000's

or do :)0-:( :D

sirpercival

should answers be deleted from the sandbox once they've been posted?

Geobits

4:00 PM

@sirpercival Generally yes.

sirpercival

okie doke

Martin Büttner

@Optimizer it's the same length as what I've got

Optimizer

1 byte shorter

Martin Büttner

was there a shorter version after ee{~{;}|}%?

Optimizer

I meant, mine is 1 byte shorter than what you have currently

(the one which does not work is 5 bytes shorter. if you do the :) :(, then its 1 byte shorter)

Martin Büttner

4:05 PM

oh okay, I was referring to Sp3000's

@Optimizer I think it's only 3 bytes shorter to begin with

Optimizer

@MartinBüttner how so ?

Martin Büttner

_,,.e&0- vs ee{1=},0f=?

Optimizer

ohhh

I was adding the {}% too :D

Martin Büttner

right...

oh I can use Sp3000's though to get rid of the {}% and save one byte

Sp3000

Mine is the same length, but I was wondering if you could save by starting with :ee

lol nvm

Martin Büttner

4:09 PM

yeah that

and then {..}f%

sirpercival

@Geobits, @Sp3000 should i post the scralphabet thing or should i give it a couple days for comments

Sp3000

If you're sandboxing at all, I'd suggest a one day minimum

sirpercival

alrighty then

Sp3000

To give other timezones at least a chance, even for the most impatient

@MartinBüttner _,,{;(},

Oh wait, empty array left

Stick a +? Same length then though :(

Martin Büttner

{_,{;(},+}%

Sp3000

4:22 PM

Oh right, implicit range

Martin Büttner

thanks :)

Nathan Merrill

given a set of items N, and each pair of items in the set has a weight M, order the set such that the sum of each adjacent items' weight is as large as possible

Martin Büttner

it's still horribly long though

CJam needs an operator for this

@aditsu ^

Nathan Merrill

is there an algorithm for the above?

Optimizer

@MartinBüttner veryyy specific use case

Sp3000

4:24 PM

You can generalise to two arrays, one values and the other truthy/falsy

Nathan Merrill

is that to me, Sp3000?

if so, what does true/false mean?

Sp3000

That was at Optimizer/Martin :)

Nathan Merrill

ok :P

Geobits

@Nathan For the sum, wouldn't it just be 2*sum(m) - m[0] - m[n-1]? Each item is being added twice except the first and last.

Zgarb

@NathanMerrill You are trying to find a maximal weight Hamiltonian path in a complete graph with weighted edges

That may be googlable

Geobits

4:28 PM

If my last comment is correct, then just put the lightest items in first/last positions and the rest doesn't matter.

Nathan Merrill

@Geobits no, that isn't right. Each pair's weight won't be summed. I need to order them such that the pairs are the greatest

@Zgarb that sounds right, thanks

Sp3000

@Geobits I think it's a matrix of pairwise weights

Nathan Merrill

yep

Longest path problem

In graph theory and theoretical computer science, the longest path problem is the problem of finding a simple path of maximum length in a given graph. A path is called simple if it does not have any repeated vertices; the length of a path may either be measured by its number of edges, or (in weighted graphs) by the sum of the weights of its edges. In contrast to the shortest path problem, which can be solved in polynomial time in graphs without negative-weight cycles, the longest path problem is NP-hard, meaning that it cannot be solved in polynomial time for arbitrary graphs unless P = NP. Stronger...

NP-hard :(

Geobits

I'm not sure what you mean now by "each adjacent items' weight". If the set is [1,2,3,4], what are you looking for?

Oh wait, each -pair- of items has a weight.

Well nvm.

That'll teach me to not read.

Zgarb

Right, Hamiltonian path reduces to it pretty directly (given graph G, construct the weighted complete graph H with same vertex set and edge weight 1 if the edge is present in G, else 0; the maximal weight of a path in H is #G - 1 if and only if G has a Hamiltonian path).

Nathan Merrill

4:35 PM

actually, this paper says that is possible because the graph is complete: mimuw.edu.pl/~rytter/MYPAPERS/fun2012_submission_10.pdf

haven't read it yet

Zgarb

@NathanMerrill Those are special kinds of weightings that are defined using trees.

"The vertices of KT are nodes of T and weight(i; j) is the distance between i; j in T."

Nathan Merrill

bah, you're right

wait...what if I take the max edge M, and recalculate each edge N to be M-N?

and then search for the shortest path?

why is that any different than shortest path?

Zgarb

The shortest path will probably not contain every node

Nathan Merrill

oh, is shortest hamiltonian path NP-hard?

New Sandboxed Posts

0

A: Sandbox for Proposed Challenges

Quaternion calculator Quaternions are extensions of the complex number system that are used in 3d graphics and some other applications. In addition to i, there is also j and k. j^2 and k^2 are both -1, however, ij=k, jk=i, ki=j, ji=-k, kj=-i, and ik=-j. Write a calculator that can calculate addi...

Zgarb

4:43 PM

Yes, because it's equivalent to heaviest path

By what you just said

Nathan Merrill

yeah, and its NP-Complete

which means that longest path is NP-Complete as well

ok, well thanks!

sirpercival

4:58 PM

@trichoplax i'm now officially getting my ass kicked by the 4DTTT ai

and it's fast

trichoplax

Are you aiming to make it a question on PPCG? A golf, or maybe a KotH?

sirpercival

@trichoplax the ai? the game? nope, it's for an app i'm making

trichoplax

Ah OK

sirpercival

even a level 10 AI (read: 10-turn lookahead) takes <1s to generate a move

there's a noticeable pause, but it's not long enough to be annoying

its strategy is interesting and not what i would have expected, but it beats the crap out of me lol

trichoplax

Sounds like a very satisfying kind of defeat

sirpercival

5:05 PM

haha yep

randomra

5:17 PM

how do you do elementwise actions with arrays in cjam? e.g. differences like [0 1 5][1 0 3]]z{~-}% == [-1 1 2]

orlp

@randomra you install Pyth

come to the dark side

with have imperative programming

Martin Büttner

@Geobits, so what do you think about a Linken challenge?

Is CJam not imperative?

@randomra .-

Geobits

Sounds fine, but I expect most answers to be a pretty straightforward graph search (depending on challenge type I guess).

I like the game though, and it has the kid's seal of approval too. He played the first 30ish levels while waiting for dinner last night.

orlp

@MartinBüttner thought it was stack-based mr. buettner

Geobits

When does it get hard, though?

Martin Büttner

5:24 PM

@orlp so? It's still imperative, as opposed to declarative or functional.

randomra

@MartinBüttner is the dot documented somewhere?

Martin Büttner

@Geobits I'm in the fourth level set now. Three are quite a few levels I need over a minute for.

Geobits

Okay, I figured they did, just curious. In the second set now and they're still pretty simple.

orlp

what game are you talking about?

Martin Büttner

@randomra it's a 0.6.5 feature, so only in Sp3000's unofficial docs

Geobits

5:25 PM

play.google.com/store/apps/details?id=com.levelind.linken

orlp

@Geobits oh I don't have an intelligent phone

Abdou Abdou

whos that agawa ?

lol anyway

randomra

@MartinBüttner found it, thanks

Abdou Abdou

i think ma sudoko was meant to be an accent but not phrasing problem

randomra

@orlp I'm learning Python, Pyth isn't that different to peak my interest :)

orlp

5:29 PM

@randomra Pyth is just the next step in the crack-cocaine world of Python :P

because most python golfs have a direct Pyth translation - which you can then further improve :P

randomra

I hope I can take away something more than reputation from golfing so I will focus on different languages.

Abdou Abdou

all these new challenges make me wish i live twice of each moment right now

orlp

@randomra don't be silly

Geobits

Saying there's more to this than reputation is bordering on blasphemy.

orlp

reputation is all that we need

Geobits

5:32 PM

(no matter how true)

orlp

it feeds us, shelters us, completes us

randomra

And Python might be useful in other places too so I give myself a chance to learn it fairly correctly before I look at Pyth at all.

orlp

To quote the famous Descartes: "I have reputation, therefore I am".

Abdou Abdou

who cares about digital reputation

randomra

HTML Framing seems to be disallowed on SourceForge so Sp3000 have to do the docs so we could use cjam like this:

Geobits

5:38 PM

Give the code section hover-tips for operators, like a bloated IDE :D

randomra

and code-completion for 2-letter instructions :D

Geobits

Hi, I'm Jammy! It looks like you're trying to bar the foo. Would you like some assistance?

Martin Büttner

6:00 PM

@randomra relevant (see bonus panel)

Geobits

That comic peeked my interest in peaking at piques.

orlp

@Geobits I stab at thee with my paki pike, I poke and poke, until I puke. Pika! Pika!

randomra

@MartinBüttner where is the bonus panel? you mean the second one?

Martin Büttner

where it says "bonus panel"

(below the two other panels)

(you need to click it)

randomra

gray/grey :D on a bad monitor is almost white, I also searched for bonus but it was an image :D

I checked to be sure if it's peak or peek, and turns out it's piqued... blog.dictionary.com/pique-peak-peek

Transcript for