« first day (4309 days earlier)      last day (539 days later) » 
00:00 - 22:0023:00 - 00:00

12:18 AM
it does work
it's what i'm doing
 
12:33 AM
nvm i changed ur answer to make it work
 
12:48 AM
@Ginger i have 10 pinned tabs
sometimes they all auto refresh at once
it can hurt
 
@UnrelatedString how many overall tabs? :3
 
in this window? 111
 
@UnrelatedString sus
 
1:25 AM
@Steffan howd u get 37 bytes julia phi?
 
1:57 AM
woah new github code search and code view looking quite nice
(an aside: it's beta access only: github.com/features/code-search)
ooh they also have code folding in file views now
left is without code viewer, right is with code viewer
 
2:16 AM
Do GolfScript actually have a chance to beat J?
 
2:32 AM
Yeah, golfscript has beaten J a few times before
 
0
A: "Hello, World!"

apollyonELF (x86/x64, Linux), 86 bytes 00000000: 7f45 4c46 b001 b20e 488d 3539 0000 003d .ELF....H.59...= 00000010: 0200 3e00 89c7 eb18 0100 0000 0500 0000 ..>............. 00000020: 1800 0000 0000 0000 1800 0000 0500 0000 ................ 00000030: 0f05 29d0 973d 3800 0100 b03c 0f05 0000 ..)..=8....<...

 
2:53 AM
@DialFrost i type code into box and it work
idk
 
wut
hint?lol
 
i just pass a specific number to BigFloat to get it to give 1000 decimal places
wait you don't even have a solution
julia has a builtin for all the constants on code.golf
MathConstants.φ in this case
@JoKing trivially?
 
@Steffan lol
@Steffan how?
i alr have 33 without putting to 1000 places
:/
 
To save you the brute-force, the precision number is 3320 IIRC
I just bruteforced numbers until it gave 1000 digits
I don't know how it works internally
 
@Steffan hehe wut
 
2:59 AM
i have no idea
i wish it just let you pass in 1000 and not 3320
 
Looks like 3320 is the number of binary digits to represent 1000 decimal digits
log(10) / log(2) ~~ 3.322
 
Interesting
 
Which if a BigFloat is...a big float, that's exactly how it'd store the number, as a specific number of bits, rather than with base 10
 
Ah, makes sense
 
Haven't been paying attention so I don't know what language this is, but might be worth checking for a BigDecimal library/builtin
 
3:02 AM
Julia
There are libraries but not builtin
 
@Steffan heres tha problem, i dont know how to apply it :P
which is why im stuck on all the math problems
welp
 
BigFloat(number,precision)
However, phi is the only one where you can just print it out, on the rest, you have to use a preciison like 3322 and then chop the last character off
 
oh
@Steffan 3320*
@Steffan ? i got 38 bytes lmao
@Steffan also why doesnt this work? print(BigFloat(MathConstants.e,3320))
 
3:36 AM
use show instead of print to save a byte
@DialFrost it depends on the hole
im using 3322 for some
@DialFrost i just explained that, also MathConstants.e can be
the precision isn't exact so you have to chop off the last character
for e, i'm using 3324 instead of 3320
 
@Steffan WAIT THATS A THING
@Steffan how lol
chop?
 
yeah
chop("$(BigFloat(...))")
It doesn't seem to be the optimal way
But idk how the optimla way works
 
@Steffan but that leaves the number as a string lol
 
yeah
so you have to use print, not show
shouldn't be a problem
 
4:11 AM
ah
tysm
@Steffan so its somehow using another way to get rid of the last char?
 
4:48 AM
I finally beat @JoKing at something in code golf yay!
 
now get a diamond
 
@JoKing :(
 
@Steffan literally just one byte
 
@JoKing btw it was lucky numbers, trivial golfs gets me 57
what method did you use for it @JoKing? Did you like remove things from a list?
 
similar to codegolf.stackexchange.com/a/45790/76162, but with del instead of L.remove
 
4:53 AM
would you mind sharing your code? I owe you hints
or at least the main part (the part where you del and do your printing)
 
The inside of the loop looks like print(r);r+=r<2;del L[r-1::r]
feel free to give me something suitably vague
 
5:08 AM
wow i forgot about the lead i had in raku for that hole
 
@JoKing lol
@JoKing did you have to set r to 0 intially?
or something
 
same as the linked answer, just for r in L:
 
you know what can you send ur whole code lmao (unless you dont want to thats also fine)
 
5:26 AM
i prefer not to post full solutions publicly, even if they're suboptimal
 
oh ok
@JoKing this should get you 57
 
5:46 AM
anyone know how to perform this in julia del l[v&~1::v or 2]?
 
 
2 hours later…
7:35 AM
36
Q: 50 digits of π in HQ0-9+-INCOMPUTABLE?

BubblerBackground HQ0-9+-INCOMPUTABLE?! is a half-joke programming language introduced in Internet Problem Solving Contest 2011, Problem H. HQ9+ is an esoteric programming language specialized for certain tasks. For example, printing “Hello, world!” or writing a quine (a program that prints itself) cou...

 
8:32 AM
@JoKing you got 57! nice
 
9:02 AM
why doesnt this work in julia? im not sure how to fix it
n=0
while n<50
print(n);n=n+2^-(n+2^n)%3
end
 
9:20 AM
Probably negative modulo behaviour / precedence
 
9:37 AM
yup i fixed it
anyone can shorten?
n=0
while n<50 println(n);global n=n+2⊻mod(-(n+2⊻n),3)end
 
 
1 hour later…
10:57 AM
 
11:12 AM
@DialFrost i assume you can just split with an empty string as the delimiter?
 
@JoKing idk
parse is also a problem
 
looks like you need to specify the type, e.g. parse.(Int64,split("124",""))
 
@JoKing ru shayan?
 
no, interesting they used basically the same example
maybe they have the same keyboard layout as me lol
 
lol
could be someone currently in this room
 
11:21 AM
Mar 27 at 9:59, by Jo King
speaking of one hand, i've been trying to design a one handed keyboard layout. How nerdy is it to make the number row the keys 1248, where you can combine them in a binary fashion for the missing numbers
this is why my examples tend to be 12480, coz that's the order the keys are in
 
hello all
I am very impressed how much faster Julia seems to be than everything else for my challenge
 
@JoKing so your design is done?
 
no, i haven't made time to work on it so i'm stuck with a sub-optimal layout
 
oh
 
11:38 AM
@JoKing 0 how?
 
that's the fifth key
i use it as a modifier for +-_= with the other keys
 
12:18 PM
So LDW?
 
@graffe I'm surprised Python is the slowest of them all
even after using @lyxal's approach
And I mean vanilla CPython
 
@UndoneStudios Interpreted languages are always going to be slow
Though your algorithm is also very naive
You should be able to match vyxal is properly optimized
Since vixal is based on python
 
12:47 PM
argh
 
1:03 PM
so pypy is compiled?
 
1:15 PM
No, I'm not sure how it's faster than C++
Probably since it uses multiprocessing
So actual speed per code is more like in the 200 range
 
pypy is magic
 
1:37 PM
@UndoneStudios that is precisely why i am trying to learn rust
 
2:01 PM
@Steffan how is 36 bytes fibonacci possible python?
 
So you're saying you suspect it's a fibbernacci
 
2:22 PM
did we like
forget LDW or something
 
Welcome to the Twelfth Language Design Workshop! The general premise is that you can post work you've done or are doing on esolangs and people will give feedback. In short, you'll get to show off our languages and their features, chat about them, get feedback, try out WIP languages, and hopefully get ideas over the next 24 hours.
5
@RadvylfPrograms
ty
 
 
1 hour later…
3:30 PM
would someone mind coming up with a little story blurb for this?
 
that links to the q
 
whoops
 
> I have neatly arranged my presents in a heap, but Billy opened the root present first! I need your help to rebuild the heap.
something to that effect
 
@Ginger Santa, in his midlife crisis, has bought a sportssled. Unfortunately it has limited cargo space. By magic there is no height limit but the footprint needs to be as small as possible to fit
 
that ones much better :p
i love how we generally dont like backstories (we, not me; i love backstories), but you cant have an cgac challenge without one
 
3:44 PM
I like backstories
Just not excessively long ones
Or when the challenge is overly complex to fit the story better
 
3:54 PM
Backstories do add some variety
But yeah people go overboard a lot and then they get in the way of the actual challenge
 
good thing I asked yall then :p
@mousetail I like this but I'm not sure it actually makes sense with the challenge
 
4:12 PM
Yea I can't really think if any in universe reason why less packages couln't be stacked on more packages
 
Because Santa's not gonna dig around for ages under a few big things to find Jimmy's new USB-C cable or whatever
 
can't santa just like summon the presents into his hand?
 
/summon Santa present
 
If santa was omnipotent the entire advent calendar would be redundant
 
Did you mean: "redundant"?
 
4:17 PM
i cent spelll
 
don't you have spellcheck?
 
Yes but it takes a while
I type way faster than it
 
4:28 PM
g++ -lgmp -lpthread jdt.cpp
/usr/bin/ld: /tmp/ccoEslnw.o: in function `calcPrimes(int)':
jdt.cpp:(.text+0x118): undefined reference to `__gmpz_init_set_str'
/usr/bin/ld: jdt.cpp:(.text+0x12c): undefined reference to `__gmpz_probab_prime_p'
 
you dont have gmp prob
 
argh stupid g++
you have to put -lgmp after the file you are compiling
 
looks like SE finally killed the HNQ on the app :/
 
F
 
theres an app?
 
4:36 PM
There was an app :P
You cannot download it anymore, and they're slowly killing it off
79
Q: Mobile app infrastructure being decommissioned

SpencerGBeginning immediately, you should know that we have begun decommissioning the remaining infrastructure connected to our mobile apps. As you may know, our mobile apps have been unavailable on the Google Play and iOS App Store for quite some time. Those of you that still had the apps installed on y...

 
hm
 
Now that HNQ is dead, notifications don't actually come through, mod stuff never worked, and it errors when I try to actually interact in any way, I think I'll delete the app soon :P
 
There are probably third party apps
 
none that have push notifications!
 
4:51 PM
ಠ_ಠ
flagged as VLQ
 
I actually think he makes a great point about tell exchangenoob stackotherwise
 
I've screenshotted that so I can remember it forever
 
I... genuinely am not sure how to handle that post
 
:o there's more
 
Please don't go investigating into a user's post history and show it publicly here
 
4:54 PM
I think he's genuinly responding to answer the question, but it's impossible to tell what his point is
 
my favorite part was where Ayer AGG'TDd'E-A went "it's elsewherestackwise tell exchangenoob stackotherwise time" and elsewherestackwise tell exchangenoob stackotherwised all over those guys
 
It's like one of the 2015 GolfScript haters fell in a paper shredder which was shoveled into a markov chain
No way this isn't a bot
 
@RadvylfPrograms Very likely
Wait is this Ginger's bot? /s
 
this is like an indonesian gambling spambot got retrained on CGCC buzzwords and then snorted a pound of coke
@mousetail no, this transcends my humor capacity
 
@Ginger Enjoy the free helpful flag, you happened to flag it as I was writing my comment :P
 
4:57 PM
Are your teenagers osmotically acquiring biases toward the PRIMARY-MAINSTREAM-WORLDSTANDARD code-look/syntax? Click here to find out!
5
@cairdcoinheringaahing :p
 
@Ginger whats vlq
 
@Seggan very low quality
I want to know what this person thinks about the popcon problem :b
 
ah
 
We don't know Ayer AGG'TDd'E-A's gender
 
@UndoneStudios As expected C++ outperforms everything else by far if optimized
 
5:04 PM
rust may be faster
also im planning to refit my kotlin to use gmp
 
Yea rust and C are on the same level
 
Rust is kinda just C++ with guardrails right
 
It's like C++ if everything is a UniquePtr
Somtimes you have to use unsafe {} for speed though in rust
 
its kinda weird
the language adverts itself as safe, but then allows you to write unsafe code just by using a keyword
 
Yeah, because sometimes Rust's safety gets in the way of performance or being able to do something in the first place
Rust makes safety the default, no language can ever make safety mandatory
 
5:11 PM
Sometimes pointer arithmetic is just the fastest way to do something, this can't possibly be made safe
 
@DialFrost it's pretty trivial
 
Since too much safety typically forces you to do weird things to evade the compiler, which still lets you make bugs, but makes them logic bugs instead of compiler-noticeable bugs
 
Rust has a lot of great features for avoiding logic bugs too though, like the enums and newtypes
Though only if you use them properly
 
@DialFrost Int.([s...]).-48 is the shortest way I think
@DialFrost Try a recursive approach
Also, (n+2⊻n)÷2%3 instead of mod
Should get you 46 bytes
 
5:32 PM
@Seggan opt-out safety isn't much less safe than mandatory safety; unsafe is basically three pop-up dialog boxes asking if you're sure you know what you're doing and a big danger tape fence around the code that's probably at fault if something goes wrong :P
 
6:09 PM
@UnrelatedString Especially since most unsafe functions have scary sounding names
 
6:59 PM
thank you MediaWiki, very cool
 
7:29 PM
LDQ: should I add a type for a never-returning function?
i.e. Kotlin's Nothing, Rust's !
 
Why not
Lets you detect unreachable code
 
7:47 PM
^
 
^
 
Not to be confused with "lets you reach undetectable code" which would be a really cool concept for a language
 
hmm, on that topic, should unreachable code be a warning or error?
i mean, error is certainly more annoying but at least it gives the programmer incentive
 
Warning
When I'm debugging code, making something unreachable with an early return or something is an easy way to skip what I don't care about
@Seggan "Incentive" to do...what? You don't need to "incentivize" the developer to fix their own code, just politely let them know it's probably broken
 
true lol
 
8:00 PM
^^
 
8:19 PM
Raku has this annoying quirk of triggering an error for valid code that looks like it might be Perl 5 code
 
...
what
how
 
for example $a=~$a is code that stringifies a variable and assigns it back to itself, but =~ is an operator in Perl, so it throws an error
you want to index into a array with ++$, e.g. @a[++$]? sorry, $] isn't a thing in Raku (why are you telling me this)
 
why
 
@JoKing does it basically say ThisIsRakuNotPerl5Error or is it just invalid syntax
 
see this 5 year old issue github.com/rakudo/rakudo/issues/1356
@Seggan it is entirely valid, it just pulls an error because it has some special case in the compiler for an operator that doesn't exist
 
8:30 PM
weird
> It's not just the variables either. I never wrote C++ in my life, yet the compiler blocks me from using perfectly valid Perl 6 code:

$ perl6 -e 'my $stuff := new class { }:'
===SORRY!=== Error while compiling -e
Unsupported use of C++ constructor syntax; in Perl 6 please use method call syntax
at -e:1
------> my $stuff := new class ⏏{ }
its basically saying "this is not c++ duh" lmao
 
Perl: Power to the programmer, who is also in our opinion too stupid to realize which language they're using
7
 
8:50 PM
> Number 2 and string "2" are 99% same, except when they aren't.
@RadvylfPrograms tbf, i have attempted to use braces in python a few times :P
 
@Seggan they don't seem to realise ~~ is a thing
(not that that helps with string numbers)
ah wait, they bring it up later, but think it's the same as =~
 
hmm, if x == 1 | 2 | 3 looks like some really nice feature i could steal for Rol
 
I've always wanted something like that
Usually in JS I resort to [1, 2, 3].includes(x)
 
i just use ors
 
Yeah I do too, for up to like two or three, but for like a dozen, not doing that
 
9:02 PM
same :P
 
@Seggan ah yes, new class [[Please eject the disc.]]
 
sadly in Rol | is bitwise or, so ill have to find some other way...
 
@Seggan how about if x in [1, 2, 3] like python?
 
hmm
 
or if x == 1 \ 2 \ 3
 
9:05 PM
@Ginger That's no different from [1, 2, 3].includes(x) in JS
It's not a special comparison syntax
 
oh right :p
 
@Ginger :eyes:
 
@Seggan or if you have a "splat" operator like in python you could do if x == *[1, 2, 3]
which would also allow other logic operators
 
@Ginger A good complerpreter would optimize that into ors though, so not a bad idea
 
yeah twas i was thinking
 
9:07 PM
@Ginger Not a bad idea honestly, \​ being unused outside of strings is kinda dumb
 
^ I think \ needs more love :p
 
rSNBATWPL uses it for fraction float literals :p
 
^^
^^^^
 
also: ~ needs uses besides not or whatever it's normally used for
I like ~, it's cool
 
@Ginger Basic uses it for integer division, which isn't too bad
 
9:09 PM
@JoKing i prefer //
 
^
// feels like / but with more... emphasis
 
as if ure really trying to floor the number
 
^
 
I prefer ÷ :p
 
ಠ_ಠ
@Seggan don't go getting any ideas now
so if you use the backslash syntax will you credit me in the credits()?
 
9:12 PM
@Ginger It's at least thematically appropriate, or if you don't like two character operators
my big problem with it is that it has a different precedence to /*
 
\​ should be the logarithm operator
 
@Ginger i wont have credits()
 
3 \ 2 is log_2(3)
 
@RadvylfPrograms just use v
 
and as a prefix for ln?
 
9:14 PM
log should be ~
 
@RadvylfPrograms haskell uses it for lambdas :P
 
so like 3 ~ 2 or 3 ~e (note the absence of a space)
 
That makes a lot of sense actually, it looks more like a lambda than anything else in ASCII
@Ginger Why the absence of a space?
Just make ~x ln(x), like JoKing said
 
@RadvylfPrograms because otherwise it could be interpreted as 2 log (some variable e)
 
Exactly
Making ~e its own operator is ambiguous and ugly
 
9:16 PM
yeah, fair
 
Just make it a unary operator, and/or have e the constant easily accessible
You're basically saying e following ~ should be special cased as its own operator, having identifiers as part of operators is really weird
 
and also please add a unary ! operator, not having that is in my top 10 most disliked python features
@RadvylfPrograms well, if the lang doesn't have implicit multiplication for identifiers separated by spaces 2 e~ is less ambiguous but looks even worse
~2 is fine
 
Or just 2 ~ E and have E imported by default
 
bitwise not and bitwise xor should be the same character, change my mind
 
@RadvylfPrograms ಠ_ಠ
I may be stupid.
 
9:19 PM
What do you mean ಠ_ಠ, how's having Math.E aliased to E in a language with a one-char logarithm operator at all weird
Ooh cursed language idea: everything is imported by default, you need to unimport libraries you don't want in scope
 
ಠ_ಠ
@RadvylfPrograms oh, I thought you meant E would just be like a module by itself so you could do import E :b
 
Well nobody said it wouldn't be :p
The leftpad approach to standard library design
 
@RadvylfPrograms finally, a language which allows me to have 200 lines of (un)?imports at the top for a reason
even worse idea: 1984, a language where all programs get sent to the developers by the interpreter and have to be approved before they can run
actually I think radvylf could have a field day designing an authentication system for that language d:
 
@RadvylfPrograms on that topic, yall think that in addition to packages that can be imported, i can have namepsaces that are required to be FQ'ed?
 
flying quiche'd?
 
9:30 PM
fully qualified
 
Like, Java-style net.minecraft.xyz?
Not a fan
Not a fan at all
 
@Seggan more details please
 
@Ginger say i have a function foo in package bar
 
mhm
 
i can write bar/foo(), or i can import bar; foo()
 
9:35 PM
uhhhhhhhhhhhhhh
what
 
but if i stuck foo in a namespace spam, i now have to either bar/spam.foo() or import bar; spam.foo()
@Ginger thats how it works now
 
I am very, very confused
you're telling me that you can import packages and then refer to the contents without the package name???
 
yeah
 
like from bar import * in python???
 
yep
 
9:37 PM
THAT IS VERY BAD
 
thats how it works in java :P
 
>:(((((
that is an awful feature
- ambiguity if two packages have identical things
- what package did this come from??? hell if I know
- autocomplete? never heard of it
- bad messy code
in summary: please do not do that for the love of PHP
 
welcome to the JVM way :P
 
@RadvylfPrograms please convince them that this plan is insane and disgusting
 
imo its stupid that you have to import a package and then still fq it to use it
 
9:40 PM
did you not read the list of reasons for that above
 
somehow java does it /shrug
oh wait, i just described my imports wrong
you have to import bar/foo; foo()
wait i forgot which slash i use
hrm maybe i should switch to :: like rust
 
just use . like a sane person
 
its too ambigous
 
somehow python does it /shrug
 
ik
but idk how to implement it
 
9:47 PM
unambiguously
 
how will it know whether foo.bar.baz is a baz in foo.bar or a baz on bar in foo
 
@graffe not g++'s fault, it's how linking works by design
 
00:00 - 22:0023:00 - 00:00

« first day (4309 days earlier)      last day (539 days later) »