« first day (2738 days earlier)      last day (2404 days later) » 
01:00 - 21:0021:00 - 00:00

01:02
@EriktheOutgolfer Sandbox link with only #<answer-id> scrolls properly for me.
Actually... isn't it like that all the time? No?
Hm.
It's really broken.
Only answers on first page works.
@Mego Any response about it?
Have you reported on mother meta?
 
5 hours later…
06:26
how many different printable characters are there in ASCII?
Or maybe I should say, nicely printable
ngn
ngn
@Anush 128 - 32 - 1
thanks
ngn
ngn
explanation: 7 bits make 2^7=128 codepoints, the first 32 are control codes, the last one is "DEL" which is usually not considered printable
thanks!
ngn
ngn
tab is among the control codes, so +1 if you'd consider that printable
06:29
I don't in this case
I am thinking of posing a fastest-code edit distance challenge. Specifically, to compute the edit distance up to distance k. It seems there is no fast code online to do this.
However, I would need to provide input strings of length 10,000+ . What's a good way to do that for PPCG?
I could give python code to produce the strings if everyone is happy with python
 
1 hour later…
07:55
@Riker out of internet? plz send me ur adress so I can send you some internet!
08:19
But how do you download more internet when you don't have any?
08:41
Man, so much for -15:
-29
A: Downvoting needs to change

Beta DecayStop the score from decreasing below -5 If the score is at -5, people should not downvote anymore, and if anyone makes the score decrease, you yourself should return the score to -5 by either upvoting or retracting your own downvote (if possible).

 
2 hours later…
10:48
@Anush tio link is the best way
11:10
0
Q: Canver string to integer once in Java

user81944At first, i can't write English very well... My question is In java, i code a code for give a number as string ex.(string x) with joptionpane,and,i made a class for convert the string to integer and show it to ke, it work... But i want to when i write x in main class, x be a integer not a strin...

11:29
@feersum 1. it's probably just severly shaped down, 2. snail mail probably? idk
12:28
Language is defined by implementation, right? 13-byte improper quine in Bash :)
@Scrooble 6 bytes
@Scrooble well, actually bash + cat
cat can be considered a library for bash
fizzbuzz is 12 bytes in bash on anarchy
how does it look?
@FrownyFrog maybe it's a pyth call :P
12:43
@FrownyFrog I think it's a gs2 call
yes, I just found the answer
it's a gs2 call
thanks
@jslip ._. Forgot about that darn $ thing
 
3 hours later…
15:30
Apparently, part of my problem is using GDC instead of DMD, because (0 0 0 0 0)(0 0 0 1 0)(0 0 0 0 1)(0 0 1 1 1)(0 0 0 0 0) takes 10 seconds on my machine, but <1s on TIO.
i find it fascinating, how twitter literally already has a mechanism for blocking nazis in Germany for legal reasons, and you can literally query the API to tell whether an account is "withheld in Germany," but having that apply everywhere was too much effort
meanwhile if you haven't sacrificed your soul to jack, your account is instantly locked the second you change your display name to contain "elon musk"
??
@dzaima I didn't do anything to the interpreter. Apparently it's just GDC (the compiler I'm using) that slowed it down :p
@Zacharý that example is still 14s though
15:46
double the input, 4x the time
I know that.
wtf
とある実装と f(x)= ―2ˣ #完全に一致では無いのだが何となく似てる
i'm not even sure how to describe this kind of hell
@Zacharý But why is the complexity of parsing an array of integer literals quadratic???
@Dennis Because I just scan postfixes >_<
15:54
Are you sure this is a practical language? :P
@Dennis After I fix a bug with and because of ulong casting, I'll work on changing that.
16:15
CMC: The Möbius function. Examples: 14 -> 1, 16 -> 0, 19 -> -1, 20 -> 0
You may use any three distinct values, rather than 1, -1 and 0 if you want
11
Q: Primitive Roots of Unity

quartataLet z be a complex number. z is an nth primitive root of unity if for a certain positive integer n and for any positive integer k < n . Challenge Write a full program or function that, given a positive integer n as input, outputs all of the nth primitive roots of unity. You may output them in ...

Me or quartata?
in response to the cmc
sorry on mobile
16:22
I don't see the connection?
just tack a sum on the end
@ASCII-only Oh I didn't know you could have long strings in TIO
@Anush I think ASCII-only means provide a TIO link to the Python code that generates the strings
Otherwise the hyperlinks would be way too long for the post
16:42
@cairdcoinheringaahing oh ok
 
1 hour later…
Okx
Okx
17:52
CMC: Round up the input, but if it is a whole number add one.
Okx
Okx
didn't think of that
Would that work for negative numbers? What is rounding up a negative number, in this CMC?
rounding up is rounding up, no matter what
it's "rounding" (integer division with 1) that might get confused
real	0m15.299s
user	0m3.425s
sys	0m11.873s
This now becomes:
real	0m0.397s
user	0m0.389s
sys	0m0.008s
It's still O(n^2) though, I just memoized every single regex.
18:00
@Zacharý Better than no improvement
That is just a result from memoizing ALL the regexes, instead of just Number.
18:35
@dzaima Look below for how my GDC runs now.
real	0m0.130s
user	0m0.122s
sys	0m0.008s
Trying to convert my Processing APL code to regular Java. Turns out 1) java's way more annoying than I thought, and 2) my project's bigger than I expected it to be already.
@dzaima Did you just figure the first one out?
@Zacharý I always have been using Processing which is way nicer
It's pushed to bitbucket, waiting on Dennis for TIO.
evaluate is still O(n^2), but it doesn't really seem to be affecting performance that much.
@Zacharý If it took you ~2hrs to speed up RAD, why didn't you do it before adding to TIO? IIRC, you kept insisting that you'd only speed it up after TIO.
18:45
@cairdcoinheringaahing I was thinking I was going to have to actually create a custom lexer/parser, but turns out matchAll, memoize, and regex make it fast when used correctly.
And that time is GDC, since it seems DMD is much faster: I can't wait to see how fast it is on TIO.
Normally, I never optimise my code, because it's never too slow, but I've given up working on Verbosity, because any changes make it take over an hour to print Hello, World! :/
What's the current amount of time to print Hello, World!?
Real time: 0.072 s
User time: 0.051 s
Sys. time: 0.017 s
TIO's timing
._.
_._
Okay, what changes are you making to get that to over an hour?!
Last thing I remember trying to change was adding for loops, but I never pushed to github and not sure where the changed code is :/
I sort of abandoned the language
> Latest commit 6c56541 on 14 Jan
18:53
RIP Verbosiy.
Ah well. I'm working on another, soon-to-be-abandoned language now! :P
I was looking forward to use Proton 2 but Hyper also abandoned it / has gone inactive.... :(
@cairdcoinheringaahing What is it this time?
@Zacharý Golfscript style language
The only new thing about it is the code page, 512 characters rather than 256
18:56
It just means that it has 3000 lines of commands, while still having a quarter of characters unimplemented :P
How will that be counted in bytes tho'?
Half the code page is two bytes
How do you distinguish it from the part of the codepage that is one byte?
Essentially, the 0xff byte is a 'continuation' byte that tells the interpreter to look at the second half of the code page for the character
@cairdcoinheringaahing ... I think I've had that idea before. You can make it even more insane by just letting the continuation go on forever.
18:59
So 0x01 0xff 0x01 encodes ÀᏤ
@Zacharý That's called Unicode :P
@cairdcoinheringaahing I mean specifically using 0xff for a 'continuation' byte (even using the same wording)
@cairdcoinheringaahing Well, how did you get the 512th one then? 0x00-0xfe, and 0xff00-0xffff, that's 511 if I'm not mistaken
@Zacharý I implemented ªº to act like ÆŒ in Jelly and am regretting allowing that many commands :P
@Zacharý 511 characters for commands, but 512 in the code page
Ah.
@cairdcoinheringaahing I might do something similar using and allowing all kinds of left operands.
Although, this will be my first language to get the It's complicated encoding on TIO :P
@cairdcoinheringaahing That's a thing?
Any others?
Technically, you could make it not part of the encoding: just have 0xff be a character that functions as a "continuation" byte, while still being the end of the codepage
Sesos is the only other
Soon it will be three, right?
@Zacharý I could, but that's more work this late in the project :P
@Zacharý Presumably
Crap ... I broke something.
19:10
@Zacharý I figured it out, I should be able to implement roughly 970 different commands, give or take 10
No, made a mistake: 1472 commands
I broke variables >_<
Correction: I broke a lot of things.
How many things?
They're fixed now.
I forgot to add a |.|\n to one of my regexes, and forgetting it broke: newlines, , {, }, [, ], (,), , , :, and certain instances of ;.
Are variables still broken?
@cairdcoinheringaahing No, that was just a result of newlines being broken.
19:22
@Mr.Xcoder How far along is Proton 2? Last time I checked it was on TIO and usable?
It was skipping right over those characters.
@cairdcoinheringaahing I feel like it's fairly incompatible with Proton 1, and lacks additional docs. Plus, I don't think it's finished completely
What, the code page?
Didn't feel like it :P
All code pages nowadays are ASCII compatible, I thought I'd shake it up :P
Why do you even need to use a codepage? It can just be straight bytes :p
19:25
So that I'm not using unprintables?
Anyone up for a game of Spyfall?
#B4#B1'No'
What does this Orst do?
I'm not up for a game rn
I'm ded' after a 20km excoursion through Austrian mountains
@Scrooble Similar to print("That doesn't look like Orst, but ¯\_(ツ)_/¯") :P
@Mr.Xcoder Summer holiday?
19:27
@cairdcoinheringaahing Doesn't look like Orst?! I quote: ᏣᏤᏥᏦᏧᏨᏔᏖᏘᎧᎿᏀᏍᏜᏮᏭ
@Mr.Xcoder Whereabouts, if you don't mind me asking?
@Scrooble That's perfectly readable :P
@cairdcoinheringaahing Europa Park in Germany and now Kötschach, Austria
in 05AB1E - Oasis, Jul 25 at 12:54, by Mr. Xcoder
Now I gtg, since I don't have much mobile data in Roaming. Been in 5 different countries in the last 3 days lol
Relatable ;))
\o
Wait, are you going now?
19:31
o\ and /o
@cairdcoinheringaahing yep but I'll be back soon though... Sorry >_<
@Mr.Xcoder Aw, that's far from where I normally go :/
@Zacharý error (works without parentheses)
also, yay
@dzaima Thanks!
Well, I broke variables (not because of the tokenizer speedup though(
19:45
@Zacharý 13 seconds
@Zacharý What should it do?
My guess is set A to 1?
Actually, It should work
@Zacharý So what does it do?
@cairdcoinheringaahing (1+A)+A←3 should set A to 3, then return 7. Right now, it value errors on A in (1+A)
@H.PWiz Dennis hasn't pulled the speedup yet.
@H.PWiz Here's the time I have now:
real	0m0.149s
user	0m0.125s
sys	0m0.024s
anyone here drink a lot of water? need an expert
19:52
o/ Sometimes
Why do you need an expert on water?
where do you Source your water from
elaborate
?
I go into my kitchen, grab a glass and turn on the tap. Fill the glass then drink
o_o I've had variables programmed wrong this whole time... I think.
19:53
Hm
But where does the tap come from
The people who installed the kitchen
guess i meant a few layers of Indirection deeper
@cairdcoinheringaahing so what happens when it runs out of water, you replace it
where do they build them
@quartata I pay the water company to give me more
@quartata That I don't know
@cairdcoinheringaahing you mean its someones job to find water for you
Thats wild
@quartata Yeah, I guess so. Presumably someone finds water for you as well
19:57
not ostensibly
Do you know the depth of the water table where you're at?
does it have to be "freshwater'
It just can't have salt
then i think the technical answer to that question, is Zero. and not just because Im not on a table
@H.PWiz ~50 meters, according to google
@quartata In that case, go find a table, lie/sit on it, then put a bowl of water under it and tell us the depth of the bowl
20:01
@cairdcoinheringaahing that sounds like a waste but ok
Not if you use the water afterwards
if we want to be picky, Were in a dish not a table. is all.
@cairdcoinheringaahing one second i need to convert the units
26
A: What are the relationships between programming languages created by PPCG users

AdámUsing the format of Webgraphviz. Links to pages about the languages can be found on TIO. # non-PPCG languages are lightgrey, non-languages are white # PPCG languages are not colored for simplicity # this has the effect of making PPCG be languages with a border digraph PPCGGenealogy { nodesep=0...

2.19 * 10^-17 meters?
its a fairly big bowl
I'm beginning to regret claiming to be an expert on water
20:07
sorry.
Nah, it's fine :P
My fault for overstating my water abilities
@Adám APL -> RAD, Dyalog APL -> RAD, Jelly -> RAD, K -> RAD, J -> RAD
@Zacharý Do you need APL -> RAD?
@Adám I'm secretly hoping that one day we will find a cycle in this graph:)
@flawr Well, in a sense. APL -> J -> Dyalog APL.
20:13
PHP -> PHP
@flawr ‽
Perl5 -> Perl6 -> Perl5 ?
@Adám ⍶⍹
@Zacharý That would be APL -> SAX -> RAD or something.
@Adám It's more like APL -> SAX -> GNU APL -> RAD or APL -> SAX -> ngn/APL -> RAD.
The APL -> RAD is unneeded though.
@Adám I hope nothing else is in that cycle =P
20:19
@Zacharý While I am generally opposed to multi-glyphs, the one good thing about ⍺⍺ and ⍵⍵ and ∇∇ is that it allows a natural extension to hyperators' self-reference ∇∇∇ and hyperands ⍺⍺⍺ and ⍵⍵⍵.
@cairdcoinheringaahing Even-higher-order functions.
This explains why hyperators exist, but doesn't provide an example
@Adám I'd hate using a tri-glyph though
@cairdcoinheringaahing that's because they don't exist (currently in Dyalog APL)
@Adám There'll be no need for hyperoperators in RAD: & and are a thing in RAD. They box and unbox functions, so you can return them from a function. ( will be the thing that replaces & from Dyalog)
20:26
E.g. ÷ is a dyadic function. is an operator which e.g. takes a dyadic function modifies it to use its arguments in swapped roles. E.g. 2÷⍨6 is 3. @ is a dyadic operator. One could devise a hyperand $ which modifies a dyadic operator to use its operands in swapped roles. E.g. 3$@- would negate at the third position (-@3 negates at the third position).
F←+&⍬
G←F+1
g←G‽
⎕←1 g 2⍝4
@Adám And presumably you could make a hyper-hyperator which swaps the arguments of $, and so on?
@Zacharý (In Dyalog, and & do two different things. launches in a parallel OS thread. & launches in a parallel APL thread in the same OS thread.)
@cairdcoinheringaahing Yes.
@Adám isn't a thing yet, is it? Or is it just undocumented?
20:30
@Zacharý It is documented, it just hasn't been connected to its symbol. Meanwhile, it is provided as a cover operator called II and the phrase ∥¨ (a common construct) is provided as a cover operator called .
Heh. That's a pretty clever choice of glyphs
@quartata To whom is that directed?
@quartata Please use reply functionality.
Ninja'd ...
|| and &
20:32
@quartata Still doesn't clarify if you're talking about Dyalog or RAD.
I think he means Dyalog, but I'm unsure.
Dyalog obviously, || is completely different in RAD
im typing on a crappy phone sorry
anyways it conveys the difference very well
@quartata OK. Yeah, we love clever glyphs.
Other glyphs I'm looking forward to are ¤ for new isolate, and — the Schrödinger operator.
probably just saved yourself a half page of documentation for the people it "clicks" with
@Adám Schrödinger operator? As in sometimes it runs and sometimes it doesn't?
20:36
(i.e. || is parallelism since they're kernel threads)
@cairdcoinheringaahing No. Rather, the result won't actually be computed until viewed/used.
(& is not)
@Adám So ... lazy evaluation?
@Adám What does new isolate mean?
@Zacharý Took the words out of my keyboard
20:37
@Zacharý Yeah, kind of.
@cairdcoinheringaahing Yeah, I've gotten pretty used to having to make things lazy with RAD and variables and side-effects.
@Adám whats the syntax on that look like
(schrodinger)
I wonder if I could've used lazy void and made my life a crap-ton easier.
@Pavel Create a new computing unit (different OS thread, even optionally on a different computer altogether) which can then be used to compute things.
@Zacharý In that case, Haskell might be a good language to do another rewrite/implementation in
20:38
@quartata fn⍰ arg is like fn arg but "lazy".
@cairdcoinheringaahing I'm afraid to implement an imperative language in a functional one.
oh so like an adverb
@Adám Does that mean things such as Haskell's infinite lists are a possibility?
dstring groupify(Segment[] args) {
	Segment[] tmp;
	foreach(i; args) {
		if(i.group == "N"d) {
			tmp ~= Segment("N"d, Value(0), "0"d);
		} else if(i.group == "F"d) {
			tmp ~= Segment("F"d, atom("+"d), "+"d);
		} else if(i.group == "M"d) {
			tmp ~= Segment("M"d, atom("/"d), "/"d);
		} else if(i.group == "D"d) {
			tmp ~= Segment("D"d, atom("∘"d), "∘"d); //FIXME: THIS COMMAND WILL ERROR ON CERTAIN INPUTS.
		} else {
			tmp ~= Segment(i.group, Value(0), i.group);
		}
	}
	return evaluate(tmp).group;
That is the hackiest thing I've written for RAD yet.
Why do all the strings end with d?
20:40
@quartata I think the APL term is Operator?
@cairdcoinheringaahing dstring (UTF-32 I believe. I was getting a ton of encoding errors when I tried to use just string)
What's the difference between that and a normal string?
See the edit
(as opposed to a C string)
20:41
Or, god forbid, a G string. (I'll see myself out)
@quartata It's an extension on the char v. wchar thing.
@Adám Could you do f←{1,f⍰⍵} ⋄ 100↑f ⍬?
@H.PWiz Does that yield numbers up to 100? My APL's a little rusty
Just 100 1s
@dzaima Fixed, but Dennis hasn't pulled yet.
20:44
@H.PWiz Oh yeah, there's no counting :/
I always forget what does
Take the first n elements of a list
(Or last, padding if needed. It also works on higher-dimensional arrays.)
@dzaima: why are is your output format with {}s? Have you not gotten to implementing a good output format yet?
@Mr.Xcoder Soon? Liar!
@Zacharý I find that output easier to read, at least while developing built-ins that deal with shapes & depths
20:50
ಠ_ಠ I have 48 trig-related commands in Orst
@cairdcoinheringaahing O_O why not just do something like for the normally unused ones, then separate ones for the basic ones
Because I don;t know what that does
(e.g. 2 2⍴⊂2 2⍴⍳4{[{[1, 2], [3, 4]}, {[1, 2], [3, 4]}], [{[1, 2], [3, 4]}, {[1, 2], [3, 4]}]})
@dzaima Ah ... I thought is was just a native format java spits out.
@cairdcoinheringaahing Variety of trig functions determined by the left argument's value.
Also, if my code doesn't have a "Complex hyperbolic arccotangent" function, why even bother? :P
20:53
@cairdcoinheringaahing Do you need separate complex functions from the non-complex ones?
I can guarantee it'll be useful when someone posts a challenge about complex trig :P
@Zacharý Well, I don't know about "need", but Python's cmath library has them, and I have plenty of commands to fill, so ¯\_(ツ)_/¯
(I mean: lump the real version and the complex version together.
Annoyingly, the cmath version returns a complex number, even for real arguments, and the math version can't handle complex arguments
I had the same problem with DMD (compiler most people use, on TIO)'s abs, yet somehow now GDC (one I just sudo apt install'd)'s.
Python's abs does work on complex numbers, but I only found that out after implementing a complex.phase command :/
20:59
Annoyingly, only sin and cos are implemented for complex numbers in D's std.complex.
01:00 - 21:0021:00 - 00:00

« first day (2738 days earlier)      last day (2404 days later) »