The Nineteenth Byte

12:53 AM

I made a few changes to the EOEIS scoring to 1) even out the scoring of the different depths, and 2) multiplying scores by a larger number so that people don't get scores less than their number of answers.

Martin Büttner

sounds reasonable

btw it's kinda funny that I went to fractran for a language where the empty program is the identity... that actually works in GolfScript ;)

PhiNotPi

It shows your... breadth of esoteric programming knowledge?

Martin Büttner

lol

if I could claim to be able to program in fractran... that would be impressive, I guess.

PhiNotPi

It's not actually that hard, or at least not as hard as it could be.

But... it can't output arbitrary numbers, so it couldn't really be used for a challenge like this.

Martin Büttner

well for this particular challenge I think it's actually trivial to find sequences that do work

like 2/1 should do

PhiNotPi

1:07 AM

Wouldn't that be an infinite loop?

Martin Büttner

wait no

yeah, I was thinking about a different form of sequence printing

PhiNotPi

Any opinions on whether I should allow different forms of "outputting a sequence", instead of returning the Nth term?

Martin Büttner

I don't think it would hurt, and it might accommodate some esolangs (if you care about that).

PhiNotPi

maybe output a continuous stream of numbers?

Or maybe just the first N numbers?

Martin Büttner

I'd probably allow those 3

(i.e. those 2, plus the one you already have)

PhiNotPi

1:29 AM

added

Martin Büttner

1:41 AM

@PhiNotPi "There must be a way to differentiate between consecutive terms" what about sequences where all numbers have the same width?

PhiNotPi

Do you have a better idea? (honestly, not sarcastically)

Martin Büttner

no, I'm (honestly) asking what your stance on those is. should they still be delimited?

i.e. do the terms have to be distinguishable with or without knowledge of the given sequence?

PhiNotPi

I think so.

Martin Büttner

It's probably easiest to always require delimiters, or you'll get into loads of annoying arguments. But the current wording seems vague enough so people might start doing this anyway.

How about "Consecutive terms must be delimited by arbitrary strings of non-digit characters."

PhiNotPi

edited

Martin Büttner

1:50 AM

I think I've got a new wallpaper. :)

Only thing is, I can't decide on the coarseness. :/

PhiNotPi

Definitely go with the 2500 so you can actually see the chords.

Martin Büttner

If you view them on 2560x1440 you can see the chords on all of them :P

I think I'll go with 10k for now.

PhiNotPi

2:18 AM

You do realize it's 3 AM in the UK, right?

#sleepspiracy

PhiNotPi

2:35 AM

I found my favorite subreddit of the day: reddit.com/r/pitchforkemporium

PhiNotPi

3:19 AM

hi

3

apparently the chat room has finally kicked the habit of starring anybody that says "hi"

New Sandboxed Posts

3:31 AM

0

A: Sandbox for Proposed Challenges

Area of an ascii polygon code-golf ascii-art geometry You should write a program or function which receives a string representing an ascii-art polygon as input and outputs ot returns the area of the polygon. The input is a string consisting of the characters _ / \ L V space and newline definin...

Doorknob

@PhiNotPi you can't tell me what to do

PhiNotPi

If you have the time, are you willing to read over the Evolution of OEIS spec?

4

A: Sandbox for Proposed Challenges

Evolution of OEIS In this challenge, the goal is to recreate the On-Line Encyclopedia of Integer Sequences one sequence at a time. Similar to the Evolution of Hello World, each answer depends on a previous answer. Over time, this challenge will create a "family tree" of the OEIS sequences. It...

I think this could be one of my most virulent challenges yet.

randomra

3:48 AM

@PhiNotPi maybe add the trick for prefix searching in OEIS

which Zgarb mentioned

otherwise looks good to me

maybe cut off some digits from numbers like 12.330684244297007 in the snippet

graph is cool!

you cant add OEIS link in the snippet, right?

Optimizer

5:50 AM

@PhiNotPi can you not show the label above or below the node, instead of inline (which makes it overlap with the curves)

Optimizer

8:00 AM

@randomra if I submit a function, do I have to print the output, or can I simply output them ?

randomra

8:56 AM

@Optimizer by output you mean return, right? "You should write a program or function ... outputs or returns the number of occurrences..." so you can just output/return it

Lembik

9:51 AM

hi all

I am thinking of relaunching codegolf.stackexchange.com/questions/41021/… with a different sort of matrix

please tell me if you think this is a terrible idea!

that is not a cyclic one

Martin Büttner

10:06 AM

That really depends on the property.

@PhiNotPi says the guy who posted the challenge with the most answers ever...

Sieg

I think I fixed the last of the bugs in my type checker. i.imgur.com/DP4gdlW.png -> i.imgur.com/mmZ8Yj9.png

Martin Büttner

10:21 AM

@Lembik "diagonally up and left"?

Martin Büttner

10:32 AM

also I feel like this is too similar, but Peter would be the better judge of that.

@randomra I like that ASCII polygon proposal. You realise though that it can be solved simply by sweeping lines?

(that's probably still very fun to golf)

Lembik

@MartinBüttner yes

@MartinBüttner it is definitely similar but also definitely not the same

@MartinBüttner it is quite possible that you can get completely different scores

Martin Büttner

that doesn't mean it's different enough to be interesting. ;) also do you know that there are non-cyclic matrices scoring higher than cyclic ones?

Lembik

@MartinBüttner just added one to the question

there is a 7 by 12 T matrix but no such 7 by 12 cyclic one

Martin Büttner

k

Lembik

we need more than one Peter :)

there was a nimrod guy at one point who seemed to be a match for Peter but he seems to have gone

oh.. what tags shall I add?

math?

combinatorics?

Martin Büttner

10:41 AM

those two and permutations

Lembik

why permutations?

( I just added it)

Martin Büttner

oh wait

no sorry, get rid of it

I still half had your Hankelable challenge in mind

Lembik

ah yes

that was tough too :)

I asked a cstheory question about that and got zero response

but then cstheory << ppcg :)

22

Q: Is Hankelability NP-hard?

I asked this question on SO on April 7 and added a bounty which has now expired but no poly time solution has been found yet. I am trying to write code to detect if a matrix is a permutation of a Hankel matrix. Here is the spec. Input: An n by n matrix M whose entries are 1 or 0. Output: A ...

cc.complexity-theory

just dumb comment questions where were already answered in the question

there are only three good *.se sites in my opinion :)

3.5 maybe

Martin Büttner

@Lembik it's on the front page of unanswered questions. it might get some attention eventually.

randomra

yeah, started with only /\_ but that was way too easy
marking closed areas with letters and asking their area might be interesting too

Martin Büttner

10:49 AM

I'm not sure what you mean by that.

randomra

that would need some row-interaction

Martin Büttner

no what do you mean by "marking closed areas with letters"

randomra

drawing

 ____________
/a  \b  / _____
\___/\_/  \___/

Martin Büttner

oh okay

Lembik

@MartinBüttner maybe... I tend to think that cstheory people don't like to say anything too new in case they could have published it :)

Martin Büttner

10:53 AM

@randomra yes that would be much harder, but at the same time it would be "just another flood fill challenge"

randomra

@MartinBüttner will think a bit about it maybe there is a smarter solution (probably isn't)

sirpercival

11:22 AM

hey folks guess what! my koth controller finally works!! :D :D :D

Sieg

@sirpercival fun fun fun youtube.com/watch?v=dJFyz73MRcg

sirpercival

@Sieg lol

here's a replay with some default bots: gfycat.com/AnnualBestBrontosaurus

Sieg

Now what's your koth about?

sirpercival

11:37 AM

I'm adding some info into the github Readme then I'll post a link

PhiNotPi

11:55 AM

@Zgarb @randomra What is the trick to searching prefixes in OEIS? I must have missed where that was mentioned.

randomra

@PhiNotPi chat.stackexchange.com/transcript/240?m=21279326#21279326 (seq:1,2,3,6,11,23 -seq:_,1,2,3,6,11,23)

PhiNotPi

thanks

I'm working on a way to change the position of the labels so that they are underneath the points.

done.

centered label text now live: meta.codegolf.stackexchange.com/a/4921/2867

Martin Büttner

looks nice

PhiNotPi

12:11 PM

maybe I could add sequence links....

(ignoring the scope creep which is going on)

Martin Büttner

@PhiNotPi and answer links? :)

trichoplax

Scope creep should be a card game. Or possibly a KotH

Martin Büttner

(link the number to the answer and OEIS ID to the sequence)

PhiNotPi

The number to the answer? Something different then the "link" on the right?

Martin Büttner

oh, in the table. I thought in the graph.

PhiNotPi

12:14 PM

I might be able to add it to the graph (table is easiest)

Martin Büttner

sure, just use the table

PhiNotPi

Hmm... it seems to replace the {{SEQUENCE}} tag in the link but not in the text. (there are two {{SEQUENCE}} but only one is replaced)

Is that normal, should I just add a second replace, or is there a replaceall that I should use?

I've gotten it to work by adding a second replace statement

Interestingly, one of the example sequence numbers we randomly picked was the unary sequence.

Martin Büttner

@PhiNotPi string replacement only replaces one. to get replaceall you need regex replacement with a /g modifier. did you switch to nderscore's suggestion that uses regex?

PhiNotPi

no

well, my plan to add links to the actual graph backfired: 42 : <a href="https://oeis.org/A012345">A012345</a>

Martin Büttner

ah okay, no worries

@PhiNotPi btw, I ended up going with 2,500 chords, but regenerated that with less opacity so the black didn't look as harsh.

PhiNotPi

12:30 PM

sequence links in table are now live

Sp3000

12:41 PM

Did anyone work out a quick way for Euclidean rhythms? Or are they just secretly saving it for when I post? :P

PhiNotPi

I have a partial solution...

randomra

@MartinBüttner floodfilling ascii-polygon seems tedious, I would probably still use a modified sweeping method. Making areas convex makes that method easier.

Sp3000

I was contemplating allowing cyclic rotations but those solutions can't win, but I think it might get confusing if there's a mix of answers...

randomra

floodfill needs to know the direction where the flood is coming from

it could be a Carcassone field score like challenge (goo.gl/LxC2xu)

PhiNotPi

If you are trying to find (8,5) for example, divide 8 into five equal parts: 0, 1.6, 3.2, 4.8, 6.4. Then you add a small constant, which I can't find easily, and which is between 0.4 and 0.6 in this case, and then truncate the number.

{0, 1.6, 3.2, 4.8, 6.4} -> {0.4, 2, 3.6, 5.2, 6.8} -> {0,2,3,5,6} -> x.xx.xx.

@Sp3000 I would say don't allow cyclic solutions at all.

Sp3000

12:53 PM

Ahaha yeah that constant...

Okay, will keep it as is then

PhiNotPi

Did you already know about said constant?

Sp3000

Well I had the divide-and-truncate part but didn't know what to do from there

Sieg

1:06 PM

Ugh. I should focus when I code. I have no idea what this does, but it works. scope[x[x2[i][0]][1]] = placeVariables(placeVariables(f[1][x2[i][0]], args[2]), args[1]);

Oh, it solves the type variables of functtion's parameters.

Lembik

@MartinBüttner it seems my definition of T matrix is causing some confusion

@MartinBüttner can you see any way I could clarify it?

Martin Büttner

@Lembik it seems like the confusion was merely caused by your typo, which you've fixed

Lembik

@MartinBüttner great.. and you pointed it out earlier to me.. thanks

at least I think you did :)

let me advertise codegolf.stackexchange.com/questions/49218/… :)

hi @Sp3000

Martin Büttner

1:38 PM

@Sp3000 that was fun :)

Sp3000

:) glad you enjoyed it - I'd tag it array-manipulation but I'm pre-empting otherwise :P

@Lembik hi

Sieg

Lol. My test failed -> I went bug hunting for 30 minutes -> noticed the test was flawed -> noticed the compiler originally gave exactly the right error

Lembik

@Sp3000 have you had a chance to look at my latest challenge by any chance?

Martin Büttner

@Sp3000 haha, yeah that's what I thought

I can't come up with a better tag unfortunately

maybe number theory or music

Sp3000

1:53 PM

@Lembik I took a look, but I won't be attempting (been busy lately). Are you sure there's no optimal single answer?

@MartinBüttner I'd put music, but the question itself doesn't really have anything to do with that... btw I demand an explanation when you have time with all the e<blah> commands you have in there :P

Martin Büttner

oh yeah I was just about to write one

Lembik

@Sp3000 As sure as I am that I can't prove goldbach's conjecture :)

@Sp3000 there is no hurry.. maybe you will find it interesting some time later

@Sp3000 I would bet some money that there isn't an optimal single answer in fact

but that doesn't mean I can't be wrong :)

Martin Büttner

@Sp3000 done

Sieg

@Sp3000 That message soon connects the world.

Sp3000

"e` - run length encode". Well. :P

Martin Büttner

2:09 PM

what? :D it gives you an array of pairs of run-length and element

PhiNotPi

I think I'm going to post EOEIS today, because I have some free time and I want to be present to make sure the thing gets off to a good start.

Martin Büttner

@PhiNotPi Already working hard on my 0-byte golfscript submission ;)

PhiNotPi

I was hoping you'd do Fractran.

Sp3000

So... now we wait for Pyth to include RLE?

Martin Büttner

@PhiNotPi sure, let me port it.

Sp3000

2:12 PM

How do we score Fractran, anyway :P

I've always had that problem

Martin Büttner

@Sp3000 in some ascii representation

but the program is empty, so...

Sp3000

I just mean in general

Martin Büttner

probably just the bytes of 32/123 54/66 87/3...

PhiNotPi

I would assume like 23/56 43/16 = 11 bytes

Martin Büttner

^

Sieg

2:14 PM

Well, that really is the minimal representation in ascii-form.

Unless you want to do shady stuff like Base64 :P

Martin Büttner

@Sieg well you could use a larger base for the numbers

ninja'd

Sp3000

Well if I ever get the excuse to update my old answers I'll fix that then

Don't want to necro bump :P

Martin Büttner

btw, there will likely be several 0-byte answers

at least oeis.org/A000027 and oeis.org/A001477 because you allow different offsets

and it's likely someone will find a language to do the zero sequence

Sp3000

I want to do the 1 sequence :P

PhiNotPi

ready to post

#YOLO

Sp3000

2:28 PM

Wait... say Martin posts 1 2 3 4 5... as level 1, then I can't post 1 1 1 1 1... as level 1 also because the next sequence would have to start with 1 in both cases and it's the same?

PhiNotPi

0

Q: Evolution of OEIS

In this challenge, the goal is to recreate the On-Line Encyclopedia of Integer Sequences one sequence at a time. Similar to the Evolution of Hello World, each answer depends on a previous answer. Over time, this challenge will create a "family tree" of the OEIS sequences. It is simple to add o...

sequence answer-chaining

@Sp3000 You can submit it as level 1, but there's no reason to (unless you like for 1 1 1 to be level 1).

Sp3000

Bad example I guess, but my question is whether or not two "The next answer should match the following terms:"s can be the same

PhiNotPi

Yes, that it okay.

Sp3000

k

Martin Büttner

@PhiNotPi are you kicking it off?

PhiNotPi

2:33 PM

right now I'm editing the snippet

Ummm.... what should I pick as my kickoff sequence?

trichoplax

0 0 0 0

Calvin's Hobbies

@PhiNotPi Props for trying a tree-based answer-chaining question. Something I've been hesitant to risk.

Martin Büttner

@PhiNotPi you could pre-empt one 0-byte sequence by doing the natural numbers

PhiNotPi

I'll do the natural numbers then.

Maybe two bytes in Element is better than 0 bytes??

Or should I do it in Perl or Java?

Martin Büttner

whatever you want

trichoplax

2:37 PM

I'm putting off learning JS so I'm likely to say silly things to waste time - feel free to remind me to learn JS

Martin Büttner

@trichoplax how are you learning it?

trichoplax

For writing Stack Snippets, does anyone have an opinion on whether it's better to focus on learning JavaScript in general, or hop straight into looking at example Snippets?

@MartinBüttner no idea yet... :)

Martin Büttner

my go-to source for any new language is learnXinYminutes

trichoplax

I need to learn it to put together my trial version of my KotH. I'm planning on starting with Calvin's Hobbies SSKotH, but I'll need a basic understanding before I can modify it

Martin Büttner

if you actually want to delve fully into JS, I can recommend JavaScript: The Good Parts

I think learnXinYminutes is good for basic understanding

Doorknob

2:41 PM

Obligatory:

6

Sp3000

As a joke I was trying to find a OEIS sequence with 1 number and recommend starting off that

But I couldn't find one in time :(

trichoplax

@MartinBüttner that's brilliant - I love that such a site exists, and a quick glance suggests it's readable so I'll start there

Martin Büttner

it's exactly what you need when you already know how to code

trichoplax

@MartinBüttner I'll take that as a compliment ;)

(but time will tell...)

PhiNotPi

The snippet seems to be working properly, although it takes a minute or so to update after every new answer.

Doorknob

2:46 PM

That would be the API delay.

Martin Büttner

yeah, that's normal

@PhiNotPi uhhh, when you said that links around language names were working, I thought you meant they are also linked in the table :D

PhiNotPi

I meant "didn't break stuff"

trichoplax

With a Stack Snippets KotH, would it be acceptable to say that only upvoted answers will be picked up by the Snippet?

Martin Büttner

@PhiNotPi that's okay too, I guess ;)

@trichoplax it might result in some tactical downvoting

Sp3000

Hah, I thought next would be ints mod 2

trichoplax

2:50 PM

Ah. Yes. I was only thinking of it encouraging upvotes to see how a bot does, but encouraging downvotes seems even more of a problem

I was hoping to automatically exclude downvoted answers but the tactical side means I almost certainly won't then

Martin Büttner

@Doorknob does GolfScript automatically convert the input to an integer? o.O

Doorknob

@MartinBüttner I'm assuming input is on the stack.

Martin Büttner

@Doorknob since when are snippets allowed? :P

Doorknob

@MartinBüttner Well, the question says it can be an argument, and that's the normal way to take arguments in Golfscript :P

Martin Büttner

GolfScript puts STDIN on the stack

but as a string

or does it also put command-line arguments there as integers?

Doorknob

2:54 PM

> The first option is to write a program or function that inputs a number (from STDIN or as an argument)

Martin Büttner

I would assume that "argument" means "command-line argument for programs or function argument for functions" :/

@PhiNotPi ? ^

Calvin's Hobbies

@trichoplax I tried that for Red Vs Blue, people didn't like it and thought it would lead to biased voting. I ended up removing the rule.

trichoplax

@Calvin'sHobbies thanks - I'll follow suit

Doorknob

If you really want, I could wrap it in {}, which would make it a "function"

PhiNotPi

@MartinBüttner Yeah, that's what I was thinking.

Martin Büttner

2:57 PM

@Doorknob or you could prepend a ~ which would make it a working program ;)

trichoplax

I'll wait until I'm feeling more controversial before thinking about a KotH with fire power proportional to upvotes

PhiNotPi

I would also like to inform you that there is very little different in score between a 2 byte and 4 byte solution.

Martin Büttner

@randomra that's depth 4

randomra

@MartinBüttner ctrl-c :/ does the snippet corrects itself?

it does :)

Martin Büttner

I don't think your submission is valid though

I don't think you can truncate sequences. the "indexing is up to you" is only how you refer to the elements I think

@PhiNotPi ^ ?

PhiNotPi

3:04 PM

Yeah... I think I mentioned that somewhere... maybe under one of the different options.

Martin Büttner

I can only find a sentence saying you can use any indexing you like, but nothing that explicitly says whether or not you can skip the first element of the sequence as given on OEIS

(which I'd find odd)

PhiNotPi

I covered that under option two, I'll copy it to option 1: "the first terms of the output must be the first terms of the OEIS entry"

Martin Büttner

ah I see

@randomra sorry :/

(I even already had a follow-up to yours)

wait

I'm being stupid

all is well

randomra

I can't follow, so what is up?

why it's invalid?

Martin Büttner

nothing is up

ignore me

PhiNotPi

3:09 PM

He thought it was invalid, but it isn't.

randomra

f(1)=first_element or f(0)=first_element would be valid too right?

Sp3000

@MartinBüttner Damnit Martin!

Martin Büttner

@randomra yes. but I somehow thought we were doing 1, 4, 18 not 0, 1, 4, 18, which would not have been valid.

randomra

i see

Sp3000

Well I think we've hit our first dead end

(two in fact)

PhiNotPi

3:14 PM

That's the beauty of a tree.

randomra

I wonder how big is the minimum maximal OEIS tree i.e. kind of cancelling out the most common routes at start (which will happen to some extent)

Martin Büttner

@Sp3000 yep

@randomra 1

there's only one sequence starting with 249 for instance

(A045254)

randomra

if you start from the empty list, like in the challenge

Sp3000

You'd have to look at how many different start numbers there are...

Martin Büttner

@randomra I don't understand. starting from the empty list, the tree is complete.

randomra

3:20 PM

@MartinBüttner what size (number of nodes) is the smallest tree in the snippet which you can't increase by a valid answer

Martin Büttner

oh, that's what you mean. I'm pretty sure that is the entire OEIS

Sieg

What is a minimum maximal tree?

I see.

I'm a slow typer atm :P

randomra

my bad, I thought you can't use 0,1,2,[3,5,7] from 0,1,[2,3,..]

so you can't use the next number which the parent had for your new number

in the real case there will be a million submissions probably :D

PhiNotPi

I am hoping for a million submissions.

randomra

you will have to submit a few 100.000 new sequence to OEIS

PhiNotPi

3:33 PM

"Contains 257082 sequences." we're over a quarter of the way there

Sp3000

Hmm Calvin's new challenge looks so ><>able

randomra

just do a 'Generating OEIS sequences' challenge :)

Sp3000

My sequence: 1. (Hint: divisors of 1)

Martin Büttner

@Sp3000 ugh, if it wasn't for the exception of the first character this would be really short in CJam

Sp3000

What's wrong with the first character?

Martin Büttner

3:38 PM

well it doesn't work like the other odd characters

I guess I could initialise the stack to a clever pair

meh

there is no real solution to a+b=1 ^ a*b=2

Sieg

Is there an unreal solution?

TheNumberOne

.5 + sqrt(7)i or .5 - sqrt(7)i

Martin Büttner

3:59 PM

@aditsu Unhandled operator after '.': ? :(

Transcript for