« first day (2591 days earlier)      last day (2252 days later) » 
00:00 - 19:0019:00 - 00:00

12:03 AM
So that explains why I did not have this problem: I only tried Windows and Linux.
 
@HyperNeutrino 1 does not è
 
I am pretty sure 1 D0ès è...
 
12:51 AM
@Pavel maybe this might help? there doesn't seem to be anything good for free though
 
1:09 AM
0
A: Sandbox for Proposed Challenges

PrincePolkaComment or not comment? Write a function which givven an all lowercase string, return the same string but with characters that are comments turned uppercase. /* causes everything until */ (or end of string) to be comment // causes everything until newline (or end of string) to be comment '\' at...

 
1:29 AM
@ASCII-only Thanks
 
anyone on linux or windows that would like to help test VSL (at least installation)?
If you could, just try to install using the (NPM) instructions here and let me know if you encounter any problems: github.com/vsl-lang/VSL/wiki/Installation
 
kk
 
k
 
@Downgoat Didn't work
npm ERR! code 128
npm ERR! Command failed: /bin/git clone -q github.com/vsl-lang/llvm-node.git /root/.npm/_cacache/tmp/git-clone-f390e774
npm ERR! fatal: could not create leading directories of '/root/.npm/_cacache/tmp/git-clone-f390e774': Permission denied
npm ERR!

npm ERR! A complete log of this run can be found in:
npm ERR!     /root/.npm/_logs/2018-03-04T01_37_22_644Z-debug.log
 
@Pavel are you executing as sudo
 
1:38 AM
@Downgoat Yes. I tried sudo and sudo -H and it's still looking in /root
I tried without sudo but that didn't work either, actually got farther though
 
@Pavel yeah, don't execute with sudo
 
@Downgoat ^^
 
@Pavel what's the error wtihout?
 
npm WARN deprecated npmconf@2.1.2: this package has been reintegrated into npm and is now out of date with respect to npm
npm WARN deprecated graceful-fs@3.0.11: please upgrade to graceful-fs 4 for compatibility with current and future versions of Node.js
npm WARN deprecated nomnom@1.6.2: Package no longer supported. Contact support@npmjs.com for more info.
npm WARN checkPermissions Missing write access to /usr/lib/node_modules
npm ERR! path /usr/lib/node_modules
npm ERR! code EACCES
npm ERR! errno -13
 
@Pavel oh RIP
 
1:42 AM
Basically, npm is shit.
 
@Pavel OS?
 
@Downgoat Fedora 27.
I have a very simillar setup to Dennis' for TIO
 
whee finally cracked jimmy23013's abstract rewriting system thing
I ended up with 10484711290952102343406842675061151801160384315778 Cs... fun
 
@Downgoat As root or no?
 
1:47 AM
@Pavel you shouldn't need sudo
 
@Downgoat Ok installed
What next?
 
now try installing with npm i -g again
 
Same error
 
@Pavel wait really try nvm use node
 
$ nvm use node
N/A: version "node -> N/A" is not yet installed.

You need to run "nvm install node" to install it before using it.
I'd do nvm install node but I just realized I gtg
 
1:58 AM
@Downgoat looking promising so far. I think. I have slow internet.
 
2:32 AM
@Downgoat It's complaining that I don't have a visual C++ compiler. I'm already installing VS, but my understanding is that installing on Linux wouldn't need a VC++ compiler, so is there any particular reason it says it needs on on windows?
@Downgoat I now have VS installed and it still isn't working.
Same error
"No Visual C++ compiler installed."
 
@Οurous wait what. can you show install log?
That is weird, it shouldn't use windows anything considering I've developed it with mac/linux in mind and windows as afterthought :P
 
@Downgoat Move to the VSL room before we fill the chat with debug logs?
 
@Οurous sure
reference for transcript: chat.stackexchange.com/rooms/46888/vsl
 
@Lembik Thanks for the paper, will read it later.
 
3:10 AM
@Downgoat It's probably looking for GCC on Unix and VC on Windows
Since Windows doesn't support GCC in a reasonable way
 
@Pavel MinGW-w64?
 
3:32 AM
@user202729 Right. MinGW isn't a reasonable way. MinGW is a pita.
 
There is nothing more suspenseful than running your submission through the test cases and hoping that it works
And nothing more tragic than watching it fail for the empty string ;(
 
...
 
4:01 AM
@EsolangingFruit Not exactly
There's nothing more suspenseful than seeing if your code (well, it's yours but you have no idea how it works) works against testcases
 
best suspenseful case for that: I submitted my code for a contest 15 seconds before the deadline (meaning I couldn't edit anything) and so if it didn't work it meant not only would I not get the points for that submission, I wouldn't be able to regain the points in a later submission lol
 
RIP
 
@HyperNeutrino >_> well why did you submit 15 seconds before deadline anyway
 
@ASCII-only Idk if u have time but since VSL now supports WASM, we need to develop bindings in both JS and VSL would you be able to help write some of these? :3
 
4:07 AM
@Downgoat hmm do you have binding example
@user202729 well, it is only 50 rep
 
But I like bounties...
 
@ASCII-only example for puts
will probably be simpler though
since I'll probably make a vslToDomString() converting VSLStr to JS Str it won't look as weird.
 
@ASCII-only because I finished it 1 minute before deadline but then I had to download to file and then the grader signed me out because of inactivity so I had to sign in again :/
I managed to get 2/15 points using the program I spent like 1 hour on
and you can actually get 2/15 points just by doing N,M,P,Q = map(int, input().split()); print(N * M - 1) ಠ_ಠ
i think
 
That's why you should take some time (30 last minutes?) to review your code.
 
I mean I couldn't get an efficient enough algorithm to solve the task ¯\_(ツ)_/¯ whatever
 
4:15 AM
@HyperNeutrino (what is the task?)
 
Problem S5 on this
basically find the MST of a graph with up to 10^10 nodes
 
@HyperNeutrino ... That's an useless problem description ...
(you know you can't read/process 10^10 nodes in a reasonable amount of time, you need to find another way)
 
ok basically you have a grid of nodes (graph-like)
there are horizontal edges and vertical edges, and you only need to store the vertical edges on the leftmost side and the horizontal edges on the top
because all rows are equivalent and all columns are equivalent
and then you have to find an MST
 
So the flights are the same on all planets, and the portals are the same for all cities?
 
4:19 AM
And what algorithm did you use?
 
anyway gtg o/
wait
I basically used a modified version of prim's
Kruskal's was a better idea but at the time I couldn't remember how to implement that
(keep in mind that I'm terrible at most graph-theory related things as of now)
 
I can easily implement a DSU but can't understand why it's n × alpha(n) ... (TODO learn)
... Wait a minute, you can simply make a horizontal DSU and a vertical DSU, it seems easy...
 
what is a DSU?
 
@HyperNeutrino ...
 
@HyperNeutrino Disjoint Set Union data structure.
 
4:21 AM
it was a dumb implementation k don't judge D:
 
(sometimes called Disjoint Set Forest, but I think DSU is easier to pronounce)
 
hm ok I have to look at what that is
also, in case you're wondering, you can't just MST horizontally and vertically and then combine trivially
 
The core data structure of the Kruskal algorithm.
And also easy to implement.
(I always implement MST with that)
 
oh okay
 
@HyperNeutrino I actually feel that you can combine trivially. Is there an online judge (right now after the contest had ended)? Or downloadable test cases?
 
give up ILP and solve this
 
lol rip
there's an unofficial online judge here: dmoj.ca/problems/?search=ccc18
the official grader isn't open to you because your school has to register for the grader
 
@HyperNeutrino (can you use code (that is written before the contest) in the contest?)
 
4:37 AM
What if you print out the code and then type in it again
 
lol I mean I don't think they really check that much whether or not you wrote the code ahead of time but you don't know the problems until you start the timer
anyway gtg o/
 
5:04 AM
@HyperNeutrino well yeah then it's like code jam, you can prepare useful functions beforehand
 
5:38 AM
@HyperNeutrino Then you can still prepare some time-consuming-to-write but commonly-used functions.
I have FFT, BIT, DSU, Point (for geometry problems), RMQ, Suffix Array, fraction, etc. (although some of them are hardly used)
 
6:05 AM
Free Borderlands 2 (Steam) if anyone wants it: J4TX3-TA6CI-QIE0D
 
6:16 AM
looks in Steam inventory
I have Portal, Ichi and Asteria as items and keys for Dex and Hacknet (and I have no idea what to do with them)
not that anyone would want those
So feel free to ping if you just want to fill up your library with random useless crap :P
 
0
Q: What is the mathematical relationship in this sequence?

canberk_gurelI am working on a path planning algorithm, which returns the "path to the goal" as a sequence of numbers. I need to find which element (row and column) of my matrix these numbers correspond to. So I am looking for a mathematical expression that depicts the relationship. The sequence is as follow...

 
 
1 hour later…
7:51 AM
@NewMainPosts Thanks HyperNeutrino for commenting and the OP for deleting their question immediately.
 
 
2 hours later…
9:46 AM
@user202729 I am very impressed Christian Seviers has implemented it!
now we just an explanation of how on earth it works
 
Anyway... I don't think I can do better.
Should I try reimplementing it in another language to see if it's faster?
(Haskell is compiled, so I'm not sure)
 
3
Q: Sort by shuffling blocks

ZgarbBlock shuffle sort The block shuffle sort is a (rather artificial) method of sorting a list. It works as follows, illustrated by an example. [6, 1, 0, 3, 2, 4, -2, -1] Break list into contiguous blocks [6][1, 0][3, 2, 4][-2, -1] Sort each ...

 
10:01 AM
@user202729 that would be awesome!
@user202729 I really don't understand the algorithm at all so an implementation in an imperative language (python, C etc) would help people work out what is going on too
as long as you don't golf it :)
haskell is about half the speed of unoptimized C I think normally
and a lot slower than carefully optimized C
even a python + pypy solution would be great though
 
I believe that this challenge can be posted now: codegolf.meta.stackexchange.com/questions/2140/…
since there is barely anything left to suggest.
 
@ManishKundu One thing: I'd change "positive" to "non-negative" since it's defined for zero too
 
@ASCII-only No it isn't. (the gamma function isn't, the factorial is)
 
@user202729 exactly?
 
16
Q: Gamma Function Golf

MegoGiven a real number t in (-10^9,13) (not including -10^9 or 13) as input, output Γ(t), also known as the Gamma function, which is defined as follows: You may not use a built-in Gamma function to solve this task, nor may you use built-in numeric or symbolic integration functions. Your output sh...

 
10:14 AM
@user202729 Since the gamma function isn't defined for all positive numbers anyway
 
@ASCII-only It is. The function is defined for all complex numbers except non-positive integers.
 
@user202729 Hmm, so that means there's one value of the factorial for any x > -1?
 
Yes.
 
does it actually count as a duplicate? :x
 
That one forbid built-ins. That was before the non-observable requirement rule.
@Mego ^^?
 
10:18 AM
oh thank god... so i can still post it right?
 
... I don't know, wait for some more time.
 
@Lembik "Sievers", not "Seviers".
I just created a mini poll under the sandbox post. Everyone vote please. (this is comment, not meta answer, so I guess polls are allowed)
 
 
3 hours later…
1:04 PM
@WeijunZhou A little late, but while they don't have any intended effect, they do cause a new chain to be created, discarding all the code before them (as do all undefined tokens, such as k or u)
 
@user202729 Nice, +3 for not a dupe.
 
Tbh I don't get the recent trend of bashing people for self-answering their challenges.
11
 
We changed a lot of things. A lot.
@WeijunZhou By learning Jelly, you know strictly more golfing languages than me. Anyway -- can be useful for debugging. But...
 
1:19 PM

 Jelly

Discussion of the Jelly programming language. (github.com/Denn...
 
I'd say that it depends on the language used. If it's something really popular, like Python, then multiple people will post solutions anyway, but if it's used by maybe 3 people (BrainFlak etc.) then it can feel a bit like cheating (?)
 
I'll respond here too if I see you message, but I didn't in this case. ( and ) are currently unimplemented.
 
the_type = type(argument)
Proof that naming things is difficult.
 
I use that a lot. If function whatever returns something best described as whatever, the resulting variable becomes the_whatever.
 
Jelly tutorial should really be "read the source code"... I still can't understand chain separator by now.
 
1:34 PM
We happen to be talking about just that in the Jelly room.
 
Yes I know.
 
Well, if you have any additional questions, just ask.
 
1:57 PM
Q: If I have low rep, and I flag as off-topic/duplicate/etc., what would happen? Review queue?
 
If I close as dupe, I have an auto comment. What happens if I flag as dupe?
 
(and your name will not be listed where the close voters' names are, since you didn't vote ot close, you only flagged for review)
 
Also is there a way to view current review queue for a particular question?
 
what? no, why would you want to do that? you can view the review history, sure, but for one particular question?
 
2:03 PM
@caird coinheringaahing @Dennis @user202729 Thank you for your replies.
 
ngn
@Lembik this comment is out of date now
 
2:36 PM
Sorry, wrong room.
 
3:04 PM
Jelly CMC: Calculate 10000000000000 ** 10000000000000 mod 1000000007. Solutions must be runnable on TIO. No hardcode. (violate 3 rules in "things to avoid when writing challenge", but this is CMC)
 
... dammit, in jelly. lol
 
Other languages are also welcome, but no power mod built-in. (now violate built-in rule, but remove language-restricted rule, as built-in and no-hardcode are the same non-observable requirement, that's only 2 rules)
 
D:
 
@user202729 You didn't say we aren't allowed to use pow and mod in the CMC
 
@Mr.Xcoder You can.
But I doubt that it can run on TIO in 60 minutes and allowed memory.
CMP: Should Jelly have no pop head and no pop tail? (i.e., equivalent to ḷ/and ṛ/)
 
imo, no.
 
Reduce by left and reduce by right work fine, as they're considered one token by the parser
 
They would eat 2 precious bytes and they both already have 2-byte equivalents.
 
But those are O(n) in the list size. (well, we don't care because itś code golf)
 
I'm sorry, but who cares?
 
3:11 PM
Also, a builtin could also be O(n)
 
If they would use Python's [0] or [-1] then I doubt that they could be O(n)
 
Are the top 2 starboard messages pinned? If so, I think it's time to unpin them both.
 
Yes they are.
 
They're not pinned anymore, pins decay after a while
 
3:14 PM
but they are right now
 
No, the star center are still white.
 
^, ^^
 
Huh. They're black for me
 
hollowed out star = pin
 
@cairdcoinheringaahing Both stars and pins are always black for me :/ They used to have a white center, but I think that changed for some reason, at least on my machine.
 
3:16 PM
CMC: output the name of a random programming language that's on TIO
 
Distribution?
 
We have a standard definition of randomness.
 
equal
or whatever it's called
 
uniform.
 
3:17 PM
Anyone want to solve my CMC?
 
that
 
Me nah
 
Q: Why does Python have a pow function and x ** y?
 
@cairdcoinheringaahing For that.
 
pow supports powmod
 
3:18 PM
(by "that" I meant my CMC)
But I forbid builtins (<-- unobservable requirement)
 
Also pow(x,y,z) computes (x^y)%z more efficiently compared to (x**y)%z
@user202729 The result should be 109787466, for reference
Compare this to this (the latter will likely not finish until the heat death of the universe)
 
hmm I have an idea for plq
 
lol I forgot plq exists
 
when is it closing? @cairdcoinheringaahing
 
3:26 PM
March 6th
 
D: I need to get to work
 
You're still way above any of us in the Robber's thread.
It's not safe until you reveal the language :)
 
We already know the winners :P
 
Dennis = cops, totallyhuman = robbers (current statistics)
2
 
3:39 PM
... I really need to read the Jelly source code.
 
@user202729 do you understand the hafnian algorithm?
or does anyone here?
I can't really read Haskell sadly
 
Ok, give up exponentiation in squaring in Jelly and switch to hafnian.
Well, I consider reading others' code and port to another language is cheating. I will read the paper myself.
(Jelly being unable to take more than 2 arguments is a huge problem)
 
But it can
Use the superscript nilads
 
Well, I still prefer working on Jelly exponentiation ... oh no I have homework (technically not, as I won't be punished if I don't complete them, but I should do) to do.
But functions still can't.
 
bounty started....
 
3:52 PM
I suggest using bounty without deadline instead.
Well, for me to read the paper?
 
@user202729 if you do that does it still show as an open bounty?
 
Consider that I have failed to get the bounty twice (ASCII hexagon chain, integer linear programming) I doubt ...
 
@user202729 I am very flexible.. I am happy to open another bounty for great answers after this one closes
and it will be twice as much :)
 
@Lembik Reply to message please. No, but TNB also has a feed for infinite bounties and you can add a notice to the question.
Oh, @user202729 solve hexagonal maze in Hexagony. do your homework.
 
@user202729 that sounds cool... I might do that next time.
 
node&the_node=n[node_i]; I guess I will adapt this variable naming scheme...
Although it seems a bit verbose. What about a_node?
Doesn't sound as natural.
 
ngn
4:38 PM
any guesses for what "shiftmult" could mean? it's from a haskell library
 
Lookup on hoogle?
(anyway I suggest reading the paper and not others' code)
 
Q: What characters should takewhile and dropwhile be?
 
Data.Vector.Bit?
 
@ngn shiftmult is defined by the user
It looks just like plain polynomial multiplication to me
 
ngn
@H.PWiz oh... I'm stupid
@H.PWiz thanks
 
4:42 PM
Well, looks like I don't have anything to do now. Still, I suggest not reading others' code.
 
ngn
@user202729 why not study smart people's code? it's a great way to learn
 
ngn
5:19 PM
that hard-coded 5 in the impl of haf bothers me...
 
5:31 PM
@user202729 (regarding your deleted answer for the transformation sequence problem) my proof only states that there is a transformation sequence if the thing about the prime factors is true, but it didn't prove if and only if. I think it's still true regardless but I didn't explicitly prove that it couldn't be done otherwise.
Though I do believe it is fairly obvious that it's an iff.
 
The paper is hard to parse
And Christian Sievers seems to have parsed it
So it's worth looking at the post parsing version too :)
 
6:07 PM
-2
Q: Play With Numbers Challenge [HackerEarth]

Harsh TrivediYou are given an array of n numbers and q queries. For each query you have to print the floor of the expected value(mean) of the subarray from L to R. Input : First line contains two integers N and Q denoting number of array elements and number of queries. Next line contains N space seperated ...

 
if anyone might know, would appreciate if you could take a look at:
0
Q: Locating crti on cross-platform basis

DowngoatMy problem is that I have a compiler which outputs LLVM bitcode/IR files (.ll/.bc) and these are compiled to object files using llc out.ll -t obj -o out.o. The problem is, is that to form an executable I (the compiler) need to link these with crti.o/crt0.o/crt1.o. Without these (even without arg...

 
Finally, Jelly has Ɗ!
 
do we not have a challenge to take the average of an array...?
 
Hmm, what?
 
is there a challenge that's just "take an array and return its average"
closest i can find is this: codegolf.stackexchange.com/q/6257/68942
 
6:13 PM
@Mr.Xcoder Oooh, that's going to be helpful :P
@HyperNeutrino Which average? Mean?
 
arithmetic mean, yes
made wiki edit for the new quicks :D
 
how did you make that wiki edit?
nothing shows up for me
 
The changes are there for me
 
cache failed me again
also >_< I hope I was able to be notified of new Jelly stuff
 
Has anyone edited in the new atom to the wiki?
 
6:23 PM
5 mins ago, by HyperNeutrino
made wiki edit for the new quicks :D
 
you mean œị? I did, yeah
 
6:35 PM
I've updated all my Jelly answers that could make use of these quicks ^_^
 
next time you should update them in small batches (say maximum 3 answers each) so as to not flood the front page :)
 
@Mr.Xcoder prefix suffix and infix?
 
@EriktheOutgolfer I wanted to flood the front page.
 
that's frowned upon...
 
^
 
6:42 PM
@Zacharý No, combine the last three links & four links, monadically and dyadically.
To be completely honest, PPCG is too inactive today to worry about the "active" page.
 
Quoi? Can I have an example of this?
 
just because PPCG is too inactive doesn't mean floods are very appreciated
 
For example, this is equivalent to this
I've updated 4 of my answers to be precise, and honestly idrc
 
What are the quicks
 
Ɗ, ɗ, Ʋ and ʋ
 
6:45 PM
Im on mobile and too lazy to go to the wiki
 
Also the behaviour of ¤$¥ has changed
 
Good to know the hooks are getting used
@Mr.Xcoder how so?
 
@Mr.Xcoder just a little detail, some of my answers are now invalid
(RIP (dyad)(nilad)(monad)$)
 
@EriktheOutgolfer they're not invalid... Just update then to say they use an older version of the language
 
@EriktheOutgolfer wait how do those behave now
 
6:48 PM
@Zacharý no, if I get such a comment I'll update to the most recent version, and would still keep my answer competing
 
Instead of len(links) > x they now use len(links) >= x+1
 
oh. what's the difference?
 
@Mr.Xcoder how is that different
 
I don't see how that changes anything since len can't return floats...
 
Actually, none I think, but the source of Jelly has changed.
 
6:50 PM
Ninjad, but it can be overloaded to return a float, I think
 
well yeah but I don't think it actually made a difference to how it behaves
@Zacharý well yes object.__len__ can be overloaded but I think links is just a regular list... idk though
 
hm, shouldn't have removed "or more" from the docs then
(cc @HyperNeutrino)
> (if not part of an LCC)
ah, so that's the difference
 
Actually, the behaviour of ¥ is the one that changed. See the commit for details
 
how so
 
drei and vier are just put under ¥, it's not within ¥ itself
 
6:53 PM
Nope
 
@EriktheOutgolfer Euh?
 
in Jelly, 53 mins ago, by Dennis
The large amount of helper links in Jelly answers annoyed me, so I've implemented Ɗ and ɗ (drei, three), and Ʋ and ʋ(vier, four) and 3-link and 4-link versions of $ and ¥.
 
'¥': attrdict(
		condition = lambda links: len(links) >= 2 and not leading_constant(links),
		quicklink = lambda links, outmost_links, index: [attrdict(
			arity = 2,
			call = lambda x, y: dyadic_chain(links, (x, y))
		)]
)
 
Ah, I was not expecting Dennis' native tongue to come up.
 
I don't think that actually changes anything in terms of behaviour
 
6:57 PM
It does, as when the non three link version of the links are used, now try could be calling the new three link version
 
00:00 - 19:0019:00 - 00:00

« first day (2591 days earlier)      last day (2252 days later) »