« first day (2857 days earlier)      last day (1992 days later) » 

12:20 AM
CMC: make a @xnor gate
 
12:43 AM
sub{shift+shift!=1}
 
@LeakyNun =
 
@Adám language?
 
it's Adám, what language do you think?
 
Anonymous
12:59 AM
Chat Mega Challenge: There are roughly 43 quintillion valid states for a 3x3x3 Rubik's Cube. What is the largest subset of valid, unsolved states that can be solved without passing through any intermediate state that is also part of that subset, using only the basic twelve moves (F, B, U, D, L, R, F', B', U', D', L', R')?
6
 
Isn't the set just the 12 basic moves applied to the identity?
Wait
Oh you mean the
 
Anonymous
@lirtosiast No, because you can use multiple moves, so long as you reach the solved state without reaching any other state in the subet
 
Lower bound: the largest of the sets of configurations whose shortest path to the identity has i moves, where 0 <= i <= 26
I'm confused, cube20.org/qtm says there is known 1 configuration requiring 26 moves and only 2 requiring 25 moves, but doesn't applying each generator to the configuration at distance 26 give you 12 such distance 25 configs?
 
1:43 AM
-1
Q: Detecting whether we are speaking about rodents or electronic mouse

NeoVeI have this task basically In this task you'll try to distinguish the meaning of words within sentences. To be more precise, you'll try to determine whether the sentence you're looking at talks about mice, the rodents, or mice, the input devices used to control a computer. I'm thinking ab...

 
@lirtosiast Actually, that could be any APL-type language.
@EriktheOutgolfer Depends on the question. ASCII-art: Neil. Mathematics-style-question: Dennis. Array-manipulation: Adám. Anything else, I'd say Dennis.
 
@Zacharý True, or many others
 
APL (any variety), J, RAD, K, I'm trying to think of some more that it would fit with.
 
 
1 hour later…
3:12 AM
CMC: output whether 4 is a prime
 
Anonymous
@LeakyNun 0 bytes is just about every programming language ever, using the truth machine version of truthy/falsey output
 
I see
 
4 is a pseudoprime for the Wilson's Theorem prime generator
 
3:33 AM
:o
I totally did not see this coming
 
4:04 AM
@LeakyNun Depends on the value of 4
 
Anonymous
It's prime for sufficiently large values of 4
 
Anonymous
Or sufficiently small
 
4:30 AM
@Mego My intuition is that the number is between 10/12 N and 11/12 N where N is the size of the Rubik's cube group
 
question: do you prefer string.substring(start, end) or string.slice(start, end) or some other keyword?
 
Anonymous
@Downgoat [:]
 
Anonymous
But if it has to be a full word for a method, slice
 
@Mego oooh yes forgot I was going to do this
 
@Downgoat are you making a language?
 
4:43 AM
yeah
does it make sense to have abs() for unsigned integers that treat it like a signed integer?
 
Anonymous
5:03 AM
@Downgoat abs should be an identity function for unsigned integers
 
@lirtosiast Or 26.
However, because 1 state can be optimally solved in reasonable time, I don't see why it is a problem. (?)
Can we prove that there exists such a set such that its complement (which contains the solved state) is connected?
 
5:44 AM
@user202729 > The positions of the cube can be separated by a concept called permutation parity into even and odd positions. Every quarter turn from an odd position yields and even position, and vice versa, so any move sequence in the quarter-turn metric ping-pongs between even and odd positions.
Every distance 26 state must be an even permutation, so every move from it must be an odd permutation, and therefore of distance 25
 
6:03 AM
is C's >> logical or arithmetic?
@ASCII-only In VSL why is subscript parsed with an array?
 
Anonymous
@Downgoat logical for unsigned, unspecified for signed but most compilers do arithmetic
 
6:23 AM
@lirtosiast Right...
Then it's probably "up to rotation and reflection".
Yes, "and their rotations".
 
someone please explain why LLVM doesn't optimize this:
true.i.i.i:                                       ; preds = %finally2.i
  tail call fastcc void @FgNprintAstringTCString(%TCString* nonnull @1) #0
  tail call void @llvm.trap() #0
to just the trap call?
 
6:44 AM
@user202729 The identity composed with a move is counted as 12 different positions for distance=1, so I don't think rotation answers my question
 
 
4 hours later…
10:52 AM
I'm wondering whether I should have if statements in my "fully modular C program" sandboxed challenge
It seems hard to make interesting test programs without it
But the spec and/or the solutions might become awfully large.
 
11:37 AM
0
Q: ASCII art H trees

nwellnhofAn H tree is a fractal tree structure that starts with a line. In each iteration, T branches are added to all endpoints. In this challenge, you have to create an ASCII representation of every second H tree level. The first level simply contains three hyphen-minus characters: --- The next leve...

 
 
1 hour later…
12:39 PM
0
Q: Pick out the subject, object, and predicate of a sentence

The Old Man and the CIntroduction This challenge will see how simple NLP libraries of different languages have grown. Challenge This is code-golf so the shortest answer wins. Given a sentence, your program should parse it with any Natural language processing library and get the subject, predicate, and object of t...

 
1:20 PM
@gnu-nobody How fast? If y>10⁹, I have no idea.
 
 
2 hours later…
3:14 PM
I am trying to rewrite x=>(y,z)=>x.indexOf(z) without = or o for a challenge.
(y,z)=>x.indexOf(z) == (y,z)=>''.indexOf.bind(x).call(y,z) == Map.call.bind(''.indexOf.bind(x)).
I think I can do it myself.
 
@user202729 is this JS?
 
Yes.
 
Without ] too.
(for the 38% off challenge)
(I still can't find out a way to do conditional loop...)
 
@user202729 dang:)
 
3:21 PM
Anyway, my idea is to repeatedly bind call to it so it ignores some first items.
Basically the core problem is to find an efficient data encoding for the challenge.
I am thinking about using a string, then map over it some way such that one type of element turns into the index and the other turns into 32.
For example: [].map.call('01010101',(a,b)=>a=='0'?32:b)
 
3:51 PM
Lambda creates a closure, which makes each instance of the array a new one.
bind doesn't.
 
4:01 PM
Challenge idea (for feedback): Choose an encoding and implement encoder and decoder for string←→nothing-array, where nothing-array is an array recursively consisting of only nothing-arrays and/or empty strings.
 
hmm
I like it @Adám
 
Unary FTW
 
@feersum Ah, true :-(
@feersum Maybe the encoding should be included in the score somehow.
 
BMO
(not necessarily a dupe, but certainly related)
 
@BMO Nicely spotted, yeah the difference is only codepoint vs character. However if the length of the encoding counts, then it is certainly new.
 
BMO
4:14 PM
Yup, and I think you should do sth. like that otherwise it would become unary as feersum mentioned.
 
@BMO Then problem of how to score and compare, and weigh code size vs output size. Test battery?
 
BMO
@Adám: It kinda depends, if the test-battery is known it would become Huffman-encoding :S
Some measure of source and enc(source)?
That could be quite interesting
 
@BMO That's interesting.
 
BMO
:D
 
@BMO However, that could be tricked by having a special case for own source and everything else verbose.
 
4:22 PM
@Adám Closely related. Encoding a string to a natural number is almos trivial
 
BMO
@Adám: Forgot about that :(
 
@LuisMendo Sure, but not if scoring is by (or including) size of encoded data
 
Ah, I hadn't seen that part
 
BMO
@Adám You could counter this a bit by measuring multiple maps of the source, such as all rotations/prefixes/substrings/permutations or whatever. It would still be vulnerable but people would need to spend bytes on that.
 
@BMO Back to Huffman :-(
 
BMO
4:27 PM
:(
 
What about a fairly large battery with random strings so no character or sequence occurs more frequently?
 
0
Q: Potentially dangerous problems?

Koala SquadI have a suggestion for a rule that should take into effect: Should questions that would require participants to produce potentially dangerous or malicious programs (intentionally or unintentionally) be automatically disallowed? I understand that it's not very difficult to find a virus to send to...

 
BMO
@Adám Then the encoding wouldn't matter much?
 
@BMO Yes it would, as simply unary would be less efficient than more clever things.
 
BMO
What about a continually changing scoring, eg. measure of all the encodings of all the submissions? That way if someone tries to special-case stuff another answer could come and change its score.
 
4:34 PM
@BMO We'd never be able to score. Although I guess a sequence where you have to encode the previous answer might work…
 
If the goal is to force the encoding to be reasonably information-dense, why not something like "sum of encodings of all strings length 10 or less"?
 
I think it's not very hard. Just catalan number and simple dynamic programming.
 
@user202729 What?
 
@Adám That one is not related to your challenge...
 
@user202729 Oh. Reply pings FTW.
 
 
1 hour later…
5:39 PM
What's a supported syntax highlighting language that has double-quote string literals, but where if and return aren't keywords?
 
@feersum Supported by what?
 
SE markdown
 
@Mego Chat Mego Challenge ftfy
2
 
@feersum Do you have a list of supported languages?
 
Ah I found one. SQL does the trick.
 
5:49 PM
@feersum What's this for?
 
Anonymous
@flawr :D
 
@Quintec I wanted to add a little color to the EBNF here.
 
0
A: Sandbox for Proposed Challenges

feersumFully Modular C: Grading code-golf c parsing You are a Computer Science professor teaching the C programming language. One principle you seek to impart to the students is modularity. Unfortunately, past classes have tended not to get the message, submitting assignments with the entire program i...

 
 
3 hours later…
8:35 PM
2
Q: Fibonacci with n numbers instead of 2

mromanHow to Given an integer n start with n ones (i.e. 4 -> 1 1 1 1). Then sum up the last n numbers, and repeat. For n = 4 this looks like this: Start with 1 1 1 1, sum up the last 4 numbers resulting in 1 1 1 1 4, then sum up the last 4 numbers resulting in 1 1 1 1 4 7, then sum up the last 4 n...

 
any recommendation for set syntax that isn't { a, b, ... }?
 
@Downgoat } a ' b ' ¨¨¨ {
sorry :)
 
O_o
 
yes!
thank you
I gave up after realising there’s no horizontally flipped ,
 
@dzaima that is quite neat!
@Downgoat what does bother you with this syntax?
 
8:43 PM
the braces are pointing seemingly wrong way and ' will probably cause issue with string and what does ¨ mean
 
@Downgoat no I mean with the original {a,b,...} :)
 
:|
 
I think {a,b,..} is probably one of the best representations as it looks just like the notation you use in math.
What are your motives to look for something else?
 
{a, b, c, ...} has ambiguity and will lead to difficulty parsing other features so just thinking about different syntaxes
 
@Downgoat what's your array syntax?
 
8:46 PM
@Downgoat so you're trying to make a new language?
 
[a, b, c, ...]
@flawr yeah
 
do you need a literal for sets in the first place?
 
would be very nice to have because I have functions like func File.init(path: String, options: Set<FileOptions>)
so wuld be much easier to do File(path: "...", options: { .read, .write }) versus let options = Set(); options.append(.read); options.append(.write); File(path: "...", options: options)
 
One thing I do think is no optimal with the math notation is that {a,a,a,a} = {a}, so even though it might seem that a set contains something multiple times, a set is actually only characterized by a function that determines whether something is or is not part of it.
 
@Downgoat if you want something that's in the middle of both, Set.of(a, b, c)
do things like ([1,2,3] [0]) + 2 work? aka do array constants only work when assigned to a specific type
 
8:49 PM
@Downgoat ah I see, you wouldn't want to use something that implies an order
 
@dzaima [1,2,3] [0] would be 1 but things like [1,2,3] + 1 could work, haven't decided whether or not to implement that
 
what about a modifier in front of the the list/array literal s[.read, .write]? (this would allow you to easily have multiple types of collections)
(similar to the 0x or 0b prefix for number literals)
 
@Downgoat ah so infers the type of [1,2,3] from the elements?
 
@flawr I was thinking about this and what I like about this is that it could allow user-define classes to have sequence literals. For that I was thinking along the lines of <Set>[1, 2, 3] but that syntax looks weird :P
@dzaima yeah
 
@Downgoat Is there a name for the characters that come in pairs? <>""´`{}[]()\/
 
8:55 PM
yeah i think flawrs prefix suggestion is the best choice. I had hoped for just overloading [a,b,c] to be possible, but it doesn't make much sense having implicit runtime conversion from [1,2,3] to a set
 
@Downgoat Why not Set([1, 2, 3]) :)
 
@flawr :| ... this might actually work— not sure if runtime overhead is optimizable though
 
@Downgoat Does your language require parenthesis() for arguments of functions?
 
yes unless it is followed by a closure/lambda e.g.: array.map { $0 + 1 }
 
@Downgoat { $0 + 1} is a lambda that adds 1 to the first argument, did I interpret that correctly?
 
8:58 PM
yup
 
@Downgoat why don't you eliminate argument-parenthesis alltogether?
 
@flawr That is a plan as part of 'command chains' but there are some parsing difficulties that we had so I've deferred them.
 
I may or may not have already asked you whether you are familiar with Haskell:)
 
a little, haven't done any major projects in it
 
are 'command chains' something similar to function composition (as in Haskell)?
 
9:05 PM
it's closer to Obj-C's 'messaging' syntax
for composition that'll be simple operator f + g though will still have to figure out how to make functions first class
 
 
1 hour later…
10:27 PM
> All my friends find it hilarious when I make jokes about vacuous truths.
 
btw: are we planning to do anything special for the upcoming 7th birthday of TNB?
 
not that I know of
btw, the 7th birthday of TNB has long passed lol
 
-- I need to think
 
how?
the room was created before 2011-01-28 :P
 
10:36 PM
It saiys the first message was on 2011-01-28
I don't think it sat there for more than 2 months without messages?
 
If the first message was on 2011-01-28, then we should prepare for the upcoming 8 th birthday
 
actually, it probably didn't even sit there for one whole day
PPCG was first launched as private beta on 2011-01-27
 
@Mr.Xcoder thanks for doing the math
not my strength
 
says the one who posted Dirichlet convolution
 
Yeah ^
 
10:40 PM
too much slithering, probably ;-)
 
numbers are just too confusing
certainly the ones greater than 3
 
yeah, who knows what 4 is...
 
The 19th Birthday will probably be celebrated :) But by that time (probably) everyone here will have long been inactive (hmm did I get the tense right ugh...?)
 
@Mr.Xcoder you probably write better than the average of those who speak english natively
I just checked the profiles of the first few guys who participated in chat, none of them seem to be active anymore (at least not on PPCG)
strangely on the "users" tab we cannot find the oldest users
the user who voted the most here on PPCG does not have any answers/challenges codegolf.stackexchange.com/users/2372/gnat?tab=topactivity
wtf
 
any undeleted answers/challenges; we can't prove he doesn't have deleted posts
 
10:54 PM
looks like he could be an SE employee
but still
 
CMQ: TNB chatter who was here in the past month with the earliest message?
 
@Adám I just saw that:)
@Adám maybe a task for TNBDE?
 
@flawr I don't know how to TNBDE.
 
@Adám do you know some SQL?
 
I did manually find that Joey was here only 60d ago.
@flawr No, I just use APL instead.
Whoa, I've actually IRL met someone who spoke here only two weeks in.
 
11:07 PM
did you know they knew TNB?
 
@flawr No. He clearly stopped coming here long before I met him.
 
can you invite him back? :)
 
@flawr I can certainly contact him. I have his email address.
 
@Adám I meant the invite function of the chat
 
@flawr Apparently. He was on SE just 5h ago. I've never used that functionality, though. Any info about it?
 
11:14 PM
I just tried it.
You can open their profile from the chat
then you see their info/stats concerning the particular room
 
@flawr Yes, I found it, but what does it do, and what is it supposed to be used for?
 
Just gives them a notification to join the room
@Adám if you want, you can leave the room and I try to invite you
 
OK, I found this. Not exactly a month, but a a month+a week. Good enough, methinks.
 
@Adám I just invited you to another room, just as an example
 
@flawr I assume you're RO there?
 
11:17 PM
hm
no you definitely do not have to be RO
you just have to be present in this room
 
@flawr Oh, so I invited Eelvex to the APL orchard. Couldn't invite to TNB.
 
strange
did you get my invites?
 
@flawr Yes, two of them.
 
there seem to be a few not-so-wellknown features here
 
11:33 PM
o/
 
I'm totally going to answer some of those ancient challenges.
 
i'm going to change my username to correspond more to other services
 
@Mendeleev But — it is elemental!
 
just letting everyone know because last time i did that, everyone freaked out (see my chat profile)
yeah it's elemental
 

« first day (2857 days earlier)      last day (1992 days later) »