« first day (3753 days earlier)      last day (44 days later) » 
00:00 - 22:0022:00 - 00:00

12:21 AM
ರ_ರ I just suggested an edit on Math.SE and autocorrect changed the edit summary from "fixed mathjax" to "fixed Mathias"
Fingers crossed the reviewers only look at the edit, not the summary :/
@user I'm sad because the song is over :P
@cairdcoinheringaahing RIP Mathias
12:36 AM
A: Sandbox for Proposed Challenges

WezlReveal by Quarters (in need of a better name) array-manipulation code-golf subsequence matrix Inspired by this: http://nolandc.com/smalljs/mouse_reveal/ (source). A valid answer: Takes a number \$w\$ and a coordinate pair \$(x, y)\$. Outputs a two-dimensional integer array with a width and heigh...

feedback please
could maybe do with another example or two but other than that looks good
@cairdcoinheringaahing Your bounty comment says "an answer" - it specifies nothing about it having to be the rickroll....
@Ausername who Neil?
@MatthewDaniels Neil
1:16 AM
@hyper-neutrino yes
Listed as cartesian_product
Underscore is important
that doesn't sound like cartesian power to me
Then no
1:30 AM
@RedwolfPrograms I hate time :/
but I may be able to attend
2:01 AM
ah yes sure
A: Sandbox for Proposed Challenges

A usernameMinecraft piston pushing In Minecraft, pistons are blocks that can move other blocks. However, they can only move at most twelve blocks, and won't work at all if any obsidian is in the way. Also, slime blocks are sticky and stick to other blocks. Your challenge Take as input a 5x5 matrix or simil...

2:22 AM
I found something funny on SE meta meta.stackexchange.com/questions/37466/…
1 hour later…
3:38 AM
Weekends are the worst :p
I know they're supposed to be break, but they almost always end up an excuse for me to do all of the chores and stuff I've been putting off for a week
Q: Challenge: compress co-occuring columns into cartesian product

wronguThis challenge is inspired by a problem I encountered recently out in the real-world, though ultimately I cheated and found a workaround. Challenge The input is a space-separated table of characters (each line is a row, and every line will have the same number of columns). Your challenge is to ...

Looks fine to me
Could probably use some cleaning up, but closing it won't do any good IMO.
well yeah it could use a lot more cleaning up
the process isn't too exactly clear
3:56 AM
A: Sandbox for Proposed Challenges

RazetimeCount my sheep code-golf matrix Given a boolean matrix representing my grass field, sheep length \$n\$ and wool thickness \$k\$, you will have to count my sheep. A sheep is a single independent unbroken line of sheep length \$n\$ and thickness \$k\$. #|##|# | # #| | #|# are all valid sheep for...

4 hours later…
7:54 AM
@Razetime it looks like a question that needs a bounty
go ahead i suppose
8:48 AM
@Ausername It's an offer designed to encourage some fun, it's not subject to "well, technically, ..." If I feel like an answer is trying to game it for the bounty, I just won't award the bounty. Which would be a shame, because it's been kinda fun watching it :P
9:04 AM
@cairdcoinheringaahing Ok, so does it just apply to the rickroll, or to anything?
9:15 AM
@Ausername I have to wait until I award my currently running bounties first lol
@lyxal I don't think whoever created SE ever thought anyone would run out of bounties :p
9:31 AM
@Ausername Only the rickroll
Oh well :P
9:49 AM
Does anyone recognise this problem? Does it have a name?
We assign each digit to either set A or set B but not both
Given an assignment we can look at all possible three digit numbers that could have been mapped to ABA, say
Let us can this number f(ABA).
Given a small collection of three digit numbers and a mapping, it might be that some of them are mapped to ABA and some to BBB say
and none to AAA
If we just look at the string of set labels where at least one of the collection of three digit numbers is mapped to we can compute sum f() over these strings
We want to make an assignment that minimise this sum
Is this a known problem?
Probably not. It seems like an extremely specific problem in a number of aspects.
How so?
It seemed a distant cousin of en.m.wikipedia.org/wiki/Assignment_problem
But too far for me to be able to make a connection
Is this base 10 too? You say "digit" which generally implies base 10.
10:05 AM
Yes. Just 0..9
Its way easier than the agent assignment problem.
Oh go on!
I can't think of an easy algorithm
I would love to hear any ideas about this
First you can assign every digit that appears in the middle to set B. (This is probably all of them, but in the worst case only 1). Then you can set up up to 8 additional sets placing each integer in its own set. For each number take it's first and last digit find the sets that contain them, remove those sets and add back their union. Since the first and last digits must be in the same set. From there you have some number of sets one of which is already labeled as your B set. Then you just brute force, all the sets are either in the A set or the B set. Worst case is that you made no
I mean in general this can be solved in O(1) since there are just 1024 cases to check and each can be checked in O(1) time.
10:25 AM
Let me try to understand this
If the strings are 101, 111, 121, 131, 141,...191 what do you end up with?
There are two equal solutions.
Sorry no, there is only one solution, everything has to be in B.
So the cost is 1000 which is the worst possible
Yeah, Its also the best possible.
I mean there is only one way to distribute these numbers so there's not really any reason to care aout the sums.
If we assign 1 to A and everything else to B we need to look at AAA and ABA with cost 1 + 9 = 10 it seems
Which is much better
You can't assign 1 to A because of the 111.
10:30 AM
Why not? 111 gets mapped to AAA
> and none to AAA
That was just a possible example
Ah ok.
@Ausername I think my punctuation is wrong though
For example if collection was 101, 121,131,141,...191 then we could just map 1 to A everything else to B and we only need to look at ABA
With cost 9
10:34 AM
@Neil Oh well, too late to change now.
@Ausername seriously :-(
@Neil wait, which bit's wrong?
In that case why not just check all 512 possibilities? It's really not a lot. I'm sure that there are further optimizations that could be done, because this problem is solvable by a lookup table, but idk the point.
I was missing a comma after the first "No"
@WheatWizard sorry for any confusion. Is it clear now?
10:35 AM
oh actually you got it right ;-)
@Neil Oh well, I had it right all along.
@WheatWizard you mean every way of making 10 digits to 2 sets?
@Anush Yeah.
10:36 AM
@Neil user's answer is wrong though
@WheatWizard that's ok but not so good with more sets and more digits
What is the formula for this that you are using?
@WheatWizard Remember when I asked this?
@Ausername yeah you ninja'd me there
And you ninja'd me just before.
@WheatWizard 2^10 I guess which becomes 4^20 if I have 4 sets and 20 "digits"
10:40 AM
Definitely not, it was 2^9 to begin with. The easiest optimization is to just account for symmetries.
I honestly was not too interested in this problem in the first place, and it is less interesting to me each time the problem completely changes.
@WheatWizard why 2^9 not 2^10?
I don't think the problem has changed has it?
Because the A and B set are identical. You can start by putting 0 in A wlog (or 9 in B or 7 in A etc.)
Symmetries sound interesting
@WheatWizard oh yes. That's cool
I guess I am looking for something that scales better than (number of sets)^(number of characters)*(cost of checking a given assignment)
More generally you can check only sets such that their order corresponds to the order of their least element. Which allows you to divide by the factorial of the number of sets.
I didn't quite understand that. What is the order of a set?
10:51 AM
It doesn't matter how you order them. One way is to just order them by the symbol used to label them, A, B etc.
Order means cardinality?
But then what is the order of the least element?
Numbers have an order?
Take the least element of each set, and then order them.
Oh I think I see
It is just a symmetry argument. Since if I have
A = {1, 3, 9}, B = {2, 5, 8}, C = {3, 4, 6, 7}
That is no different from:
A = {3, 4, 6, 7}, B = {2, 5, 8}, C = {1, 3, 9}
So we choose an arbitrary but consistent way to order them. That is A is always the set with the smallest minimum, C is always the set with the largest minimum. Then only the first ordering is valid and the other 5 are not checked.
11:32 AM
That is good point
12:05 PM
A: Sandbox for Proposed Challenges

venusMeta game theory king-of-the-hill javascript The bots must choose to cooperate or defect before they know what game they are playing. Rules In each game, 2 bots will face off against each-other. A game consists of 100 rounds. Each bot will play 1000 games against every other bot, including itself...

1 hour later…
1:28 PM
@WheatWizard I wonder if I can make an interesting challenge from this
From the entire thing or that comment specifically?
The entire thing. If 10 is 100 it's not so easy
I don't think it is very interesting.
Because you can't see the motivation or another reason?
It's a tricky optimization problem
I don't think it is a very interesting optimization problem.
1:35 PM
:( each one to their own, said the woman as she kissed the cow
But I would love to know if it's an instance of something well known
I find it interesting
@Iamaperson hello!
Hello! I just learned what Code Golf is
I am actually from math.SE
I was just randomly checking some chat rooms and I found this one
@Anush your parent user is also not from Code Golf
Wait @Anush you also have an account on math.SE I just saw!
2:32 PM
sup everyone
hello o/
@hyper-neutrino hi
2:46 PM
Hi :)
3:20 PM
Imagine my surprise when I saw these questions apparently being asked on Workplace.SE
wait, what lol
idk why but Worldbuilding.SE now has that logo, and Workplace has a 'W'
still normal on my end
userscript bug?
interesting. can't replicate tho
No, I added SO's dark mode CSS to all SE sites using Stylus
3:24 PM
if workplace has the writing logo and worldbuilding has the workplace logo
...it couldn't be some kind of shift in lexicographic order
I'm trying to find the answer I copied the script to download the CSS from so I can update it, but it seems to have disappeared :\
the real challenge is coming up with a good next line
still, I think I'll stop at day 12, before it becomes unlucky
Which almost nobody did, unforunately
@Neil But Day 12 is the 13th day, so you can't stop then
sure, 1-indexists have to stop a day earlier, but I'm not limited by that
> 1-indexists
4:02 PM
When are peak hours for this chat room on a Sunday?
4:14 PM
there are none
codegolf.stackexchange.com/a/225323/80214 I want to know what this guy's head is made of
jelly obviously
lol, love the way he's written something to convert from fen notation as well
almost every single time he answers in Jelly it's some insane complicated problem with an enormous train
doctors are built different
ok nick kennedy is just way too OP at long jelly answers lol
@Razetime I guess inflammatory bowel disease changes you
4:51 PM
I'm just at awe of this guy
They even have a 418 byte answer
418 bytes of jelly
Don't encourage these kind of questions because there may be a chance of asking I would like to play video using R. — Avinash Raj Aug 7 '15 at 11:32
2 hours later…
6:56 PM
@Razetime Ooh... I want to take a shot at that one... but the "if it's a cycle remove it" rule looks hard to implement.
@Razetime ugh, I totally forgot to have a go at answering that question :-(
@Neil Dew it
1 hour later…
8:11 PM
Well, today's the last day of having Arial on CGCC (and other sites)
Going to take a bunch of screenshots so I have something to compare with if I ever want to make a userscript to revert it
make a userscript to set it all to comic sans
Nah, wingdings :P
For the record, I think this is yet another silly move by SE.
I'm 100% going to need to make a userscript, it looks awful on Ubuntu
People say comic sans is bad...
8:19 PM
The font Ubuntu looks awful on Ubuntu?
It looks horrible anywhere
I thought it was designed for it
It's like a cross between comic sans and the default sans-serif font
I don't know, I quite like it
8:22 PM
@RedwolfPrograms I do not like this
I like this
Roboto looks fine, but not everyone has the privilege of using a device with a good default system UI :p
We should switch laptops :P
I just discovered the Go font today. New favourite humanist typeface for me.
CMP: Favorite font? Mine's Droid Sans Mono
8:23 PM
@user I agree that typefaces are important for one's "look", but then the site should be consistent across platforms.
@user Depends on the purpose. Reading a paper book? Reading online docs? Code?
Roboto for sans serif, unsure on serif, probably Roboto Mono or the Github one for monospace
@Adám Any or all of them, I guess
@Adám Mm, but different platforms probably deal with different fonts better (at least according to the MSE question)
@user Not so if you server the fonts.
@user Not sure that's how fonts work lol
Oh ok
I'm not a web developer :P
8:26 PM
Well, I'm pretty sure you aren't a spider
You are a normal human typing with your human hands
If you've never had to try to make a website that looks nice using IE compatible HTML and CSS, you are probably a lot happier than me
@cairdcoinheringaahing Was gonna link exactly that :P
@RedwolfPrograms Why are you worrying about IE, though? Even Microsoft recommends Edge now
I don't anymore
But it's still important to support for some stuff
Lots of companies and governments are still running XP
In ten years or so maybe we'll all have moved on to reasonable auto-updating browsers and web development will be a pleasant experience
@RedwolfPrograms ಠ_ಠ
@RedwolfPrograms Righhhht
Because obviously we won't have technologies becoming obsolete and people who continue to use them
All of the major browsers and OSes are auto-updating I think, which should help to prevent a lot of that
8:33 PM
so i heard there's some event coming here.
I don't know how code golf works. can you explain to me? :)
8:34 PM
You try to write the shortest code you can to solve a problem
oh, okay.
@user Selectric for code. Go for humanist sans-serif. Atkinson Hyperlegible got grotesque. Probably Garamond for books.
@MatthewDaniels Try answering some questions to get the hang of it. You might want to try this factorial challenge
so if I say "write a program that outputs "Never gonna... nevermind", i just make some program then try to shorten it?
i might go there soon. right now, I am going site after site.
then sometime, I can go back to Stack Overflow.
wow, they're clever. looks quite weird, ngl.
@Adám The Go font looks cool, Atkinson Hyperlegible doesn't look bad either
@MatthewDaniels Yeah, since they're supposed to be as short as possible, answers can get pretty odd-looking
8:39 PM
well, nice talk. bye. :)
There are also a few languages there that were made specifically for golfing, but don't let that discourage you
I'll go somewhere else first.
9:08 PM
Is this worth writing up as a challenge? Given a list of polynomials and an integer x, compose all the polynomials together and evaluate at point x? It's similar to this, but extended to any number of polynomials and involves evaluating the final composed polynomial
imo, no, composing multiple doesn't add much over composing two, and evaluating isn't that interesting
Hm. I feel like it'd just boil down to "reduce the list over the function in the answer to that challenge" and then just "plug in a value". Though, in that case you don't actually need to compose the polynomials, so it might lead to some other approaches?
Actually this is just "apply a polynomial" but multiple times
you don't actually compose, you just apply to x and then apply the next to that result, etc.
Well, I have a 4 byte Jelly solution, which doesn't work for that challenge
@cairdcoinheringaahing is there any difference between evaluating the composed polynomial and evaluating the polynomials one by one (feeding back the results)?
it's just apply all the polynomials to the input
so even less interesting than actually composing them
9:12 PM
In which case, CMC: Given a list of polynomials and an integer x, compose and evaluate the polynomials at x :P
4 bytes in Jelly to beat
So like for and 2x-1, with x being 4, the answer'd be 49?
3 bytes in jelly i think
can't find a polynomial eval builtin in jelly
It's there, but it isn't called that :P
9:14 PM
yeah, I know ⊥ in APL does it but I don't know what it is in jelly
@hyper-neutrino You have to reverse the list before reducing, as composition works "out" from the center, not in
x^2 + 3x + 5, x + 1, x = 3 should output 33
Far from the shortest possible, but p=>x=>p.reduceRight((n,d)=>d.map((r,i)=>n**(d.length-i-1)+r).reduce((r,t)=>r+t,0),x)
@cairdcoinheringaahing wdym out from the center
Ah, ḅ
9:16 PM
@hyper-neutrino Like in f(g(h(x))), you eval h, then g, then f
oh do we have to take them in LTR order?
i'm taking that as [x, h, g, f]
That seems like an iffy I/O format tbh
well then 4 bytes i guess with a trivial modification :P
LTR vs RTL is typically up for debate, but saying "I take a list and a value as the first element of the list" seems... no
Taking the polynomials in LTR order on the left and x on the right, 4 bytes
or if that allows RTL then it's trivial 3 bytes
9:19 PM
⊥⍨/ in APL if I can take it in HN's order
@hyper-neutrino What a surprise, that's what I had :P ƒ is a surprisingly good addition
@rak1507 ey nice
same approach :D
yeah well no surprise there :P
in fact it happens that my solution is a direct transliteration of yours :P
more like mine is a direct translation of yours, but sure
9:20 PM
well same thing :p
remind me to make my base conversion attempt to determine which side is the base and which side is the digits so you don't need the @
@ngn shame // doesn't work
@rak1507 yeah, (/)/ doesn't either. / is an adverb.
Is it just me, or have we really not had many new challenges in a while? Of the last 10 questions (going back 4 days): 2 were by me, 1 was by Bubbler, 1 by Wasif and the rest were closed/spam/dupes :(
9:21 PM
@cairdcoinheringaahing yeah, not just you, there are a few sandboxed ones that have been there for ages
would be nice to give the owners a nudge
9:38 PM
Ok, so going by the newest, open questions. This month, we've had: 6 on the 1st, 5 on the 3rd, 2 on the 4th, 1 on the 5th, 2 on the 7th, 1 on the 8th and 1 today
I've posted 16.7% of those tho, which is probably why I feel like there's not that many new challenges
@cairdcoinheringaahing i think challenge authors should focus more on quality than quantity
Agreed, but the two aren't mutually exclusive. We normally get at least 1 good challenge a day, but recently, we've been getting more off-topic/spam than good questions. We've had days recently where the only things posted were spam/off-topic posts
Q: Implement cowsay

MakonedeIntroduction cowsay is a Unix command made by Tony Monroe, written in Perl. It outputs a picture of a cow saying given text. For example: ________________________ < Typical cowsay output! > ------------------------ \ ^__^ \ (oo)\_______ (__)\ )\/\ ...

@RedwolfPrograms New Posts : have the usernames be links to the user's page
It's consistent with pretty much everything else in SE and saves a few clicks for finding out about the user
how much rep do you get per upvote on a question?
damn, questions are way too op for rep
Nah, I prefer 10 either way. 5 for questions felt like a punishment, especially on a challenge-centric site
interesting that you view this site as challenge centric, I view it as answer centric
9:53 PM
I'd argue it's equally both and thus it makes sense for them to be equal in value
fair enough
in general it's 5 / 10, CGCC is just different (idk if you knew this, just felt like mentioning in case you didn't)
@rak1507 Personally, I find far more challenges interesting than I do answers
@hyper-neutrino didn't know this, thanks
@hyper-neutrino No it's not
9:54 PM
wait what
@cairdcoinheringaahing I'm the opposite, challenges are boring, answers are interesting
10 for questions is network wide
did it get reworked across the network
lol, what
okay then
At least 90% of answers on the site just aren't interesting. They're very often just ports of other, more interesting answers into whatever language the answerer likes, or trivial implementations of some formula in a (not very interesting) challenge
And that's fine. A large portion of my answers are that. Doesn't make them interesting tho
yeah, most answers are boring
not going to argue with you there
but I'd say most questions have at least 1 answer on them that's more interesting than the question
9:57 PM
hi. :)
welcome back
Even worse is when someone finds a really interesting answer, and then it's ruined because every single other answer is "ports X's answer, go upvote that"
@MatthewDaniels Hello :)
@cairdcoinheringaahing yep
just found out that Code Golf SE is in a deep spot from the data list.
Which data list?
9:59 PM
just go to data.stackexchange.com and you'll know what I mean.
Oh, SEDE? Yeah, we aren't the most active site on the network :P
We have the highest answer:question ratio tho :P
00:00 - 22:0022:00 - 00:00

« first day (3753 days earlier)      last day (44 days later) »