« first day (4179 days earlier)      last day (666 days later) » 

1:02 AM
CMP: should I make bitwise or logical operators monographs? i cant afford both
 
@Seggan bitwise
 
yes
thank you
 
^^^
 
It's probably been there for a week now
I just haven't checked since lol
@Seggan you can simulate logical or with addition and logical and with multiplication if need be. The use of dedicated logical comparisons would be for short circuit stuff
 
python-style logical can be pretty useful if it vectorizes, depending on how it vectorizes
but bitwise are also pretty useful
i would probably lean logical
but it's close
 
1:11 AM
0
A: Sandbox for Proposed Challenges

tybocopperkettleSimplify Rubik's Cube Moves Background On a Rubik's cube there are 54 moves that you can execute, for example, turn the right face anti-clockwise, or rotate the top face and the horizontal slice twice. To notate any move, each face (or slice) has a letter assigned to it. To move that face clockwi...

 
1:39 AM
Thought this was parenting for a second
 
@Catija Could you please unfreeze Mathematical Regexes?
 
 
3 hours later…
4:32 AM
you see I couldn't have seen @NoHaxJustRadvylf's comment, but then i had a very good idea, I used F5, see using F5 gave me a whole new perspective and i was able to see his comment
 
in CG&CC-gaming, 3 hours ago, by lyxal
Also, I was watching a video on escape rooms in minecraft and then I had a very good idea. I used f5. See, using f5 gave me a whole new perspective and I was able to see a Radvylf comment making an f5 joke
Beat you to it by 3 hours :p
 
Radvylf is everywhere
(on the videos Lyxal watches at least)
Plot twist: Lyxal is Radvylf('s (clone|cousin|brother|dad))
 
4:53 AM
man, i want to write up a "classify japanese verbs into class 5 or 1" challenge similar to Distinguish between Masculine and Feminine Nouns in French within 100 characters but i can't find any plaintext lists of japanese verbs at all
best i've gotten are wiktionary category pages type 1/class 5 and type 2/class 1 but you need to paginate through to get them
the wiktionary lists are also in kanji, which i don't want because i only want the verbs in hiragana form
(if i can get them in romanji form that would be ok as well)
 
@lyxal lol
 
5:40 AM
@emanresuA you're telling me... people with a shared interest in code golf... have shared interests??
 
Lmao
 
5:58 AM
@pxeger probably...
 
att
6:11 AM
@bigyihsuan how would you deal with e.g. 要る/居る?
@pxeger nah, couldn't be
 
 
4 hours later…
10:14 AM
is there a challenge to find the minimal multisubset of which the multiset input is an exact multiple? e.g. aabbab -> ab * 3?
(and if not, would it be interesting?)
 
What would the output for baab be?
 
I assume these are unordered sets, right?
 
13
Q: Writing erasable code

Wheat WizardMy phone number (which I will not be sharing here) has a neat property where there is a two digit number, which when iteratively removed from my phone number will eventually remove all the digits. For example if my phone number were abaababbab Then by repeatedly removing ab we would eventually ...

This fits the description, but it's not code-golf.
It also might not be the challenge you have in mind.
 
10:37 AM
Sometimes, Jelly's design choices confuse me. For example, œṡ (the "split the left argument at the first occurrence of the right argument" atom), for some reason, vectorises when the right argument is a list: Try it online!
 
That kinda makes sense. Sounds like it's meant for splitting on simple values
Well okay it would make sense if strings weren't lists
 
@lyxal Yeah, the normal split atom, only splits on "simple" types. But, it doesn't vectorise, it just does nothing if the right argument isn't "in" the left argument
 
After looking at the different behaviours closer, I can safely confirm that I haven't a single frick what Dennis was on when making it vectorise
@Zionmyceliaadamancy I said that forgetting that strings aren't "simple" values :p
 
 
2 hours later…
1:06 PM
yeah, they are unordered sets
@WheatWizard not really, my mistake for oversimplifying the example
 
 
1 hour later…
2:23 PM
@Zionmyceliaadamancy i can't remember the last time i've wanted to vectorize the thing being split around
how could this happen
honestly jelly has a lot of issues handling lists intelligently; i comes to mind
 
3:15 PM
@att i'm wanting the test set to be hiragana only, so no katakana nor kanji in the test cases

test case format would be `VERB CATEGORY` so 要る/居る would have test case `いる 5` or `いる 1`. verbs that can be either, i'm allowing submissions to also output either (as long as it's consistent)
 
 
3 hours later…
6:21 PM
@bigyihsuan you can use AWB
22 hours ago, by Leaky Nun
CMC: Given one lowercase letter (a-z) output the corresponding letter from z-a (a -> z, b -> y, ..., z -> a)
repost ^
 
Vyxal, 6 bytes: kaḟkzi
I believe it can't get any shorter than that (in Vyxal)
 
Fig, theoretical (unimplemented yet), 5.762 bytes: i$calca
 
ugh
Can't believe you theoretically beat me by 0.238 bytes!
3
 
6:38 PM
another 6-byter: kaḂ?ḟi
 
if i had one of those reversed alphabet constants i could shave off another .8 bytes
speaking of which, i should probably go though vyxal's constant lists and steal them lol
 
hmm bifurcate
 
nahh i was looking at that kz in the first submission
 
Theoretical 5 bytes, but doesn't work due to a bug: C⁺vεC
Python, 24 bytes: lambda x:chr(219-ord(x))
We can use the old Vyxal and it will work though: 5 bytes
 
noooo
 
6:46 PM
:^)
 
adds cz to the spec
 
You'll have to add ord and chr to the spec
 
i have that already
 
And you'll also need something like
Index of next char in codepage
So that 219 is two bytes
Then you can beat
 
nah i am referring to my idea of a reversed alphabet constant
then itd be 4.939 bytes
 
6:48 PM
Yeah ig that's still less than 5
starts adding new stuff to Chocolate for this CMC
 
@Steffan woah nice
didn't think of that
 
I reported the bug with and modifier
 
Actually, might that make a good challenge on main?
 
IMO it's a little too simple
 
^^
i dont think so
 
6:50 PM
Also, I'm not sure that it's not a duplicate? idk
 
we have plenty of simple challenges
 
And most are bad
We kind of need some, for esolangs to participate in and for convenient dupe targets, but simple challenges should be avoided
 
i'm reasonably sure it does exist on main
57
Q: Swap the Alphabet

DowngoatIn this challenge, you will be "reversing" the alphabet or swapping a-z with z-a. This is commonly known as the Atbash cypher. Because this transformation makes the output look like some foreign language, your code will need to be as short as possible. Examples abcdefghijklmnopqrstuvwxyz zyx...

 
I thought so
 
6:51 PM
though it does require handling of case and non-alphabetic characters
that's not a very large difference
 
Ahah, another 5-byter: kakzĿ
 
still 4.9 bytes :P
 
@Neil reminiscent of codegolf.stackexchange.com/q/215858/85334 but i don't think a direct multiset version exists
it probably would be interesting
 
@Steffan the heck why does Ŀ even exist
 
it can actually be useful sometimes
ive seen it in a few posts
 
6:59 PM
@Steffan you got ninja'd by Aaron lol
 
7:17 PM
Chocolate, 5 bytes: ∂cacz
Transliterate is very very useful
I use it constantly
Even Python has it
And bash and ruby and etc
 
wait what
never heard of it in a real lang
 
tr in ruby and bash/zsh, translate in python
 
wow
ok time to add a P overload
 
replacing things with other things is useful
 
what does translate even do
 
7:25 PM
Zsh, 38 bytes: Try it online!
 
@AidenChow replace letters in a string using a mapping between 2 other strings
 
ive seen ppl just input strings into translate, like how does that work
like x.translate(some_string)
without the maketrans part
 
lemme check
actually zsh 36 bytes: tr a-z zyxwvutsrqponmlkjihgfedcba<&0
 
7:31 PM
basically because you know you can index into a string just like a dict
so if you have a dict {46: "h"}, you could instead have a string where the 46th index is h
based on codepoints
so here's a python thing using that hack: 57 bytes
 
7:49 PM
@LeakyNun AWB? seems to be what i might need, i'll check it out
 
8:02 PM
@mathcat kaḂĿ
 
noooooo
 
mwahaha
 
though with transliterate i could do Pcacz for 4.116 bytes
 
Still winning ;)
 
@Steffan you don't need the <&0. (That's basically a no-op unless it's the only thing in the command)
 
8:16 PM
for 6 bytes, O219_Ọ
beats ØaiCịØa which it would tie if i and didn't have opposite argument orders
incidentally i feel like this could end up being 3 bytes in perhaps with some design decisions i've been thinking of
namely, list nilads implicitly indexing into themselves in cases where it's not useful for them to act as nilads--both "unparseable" chain positions, and possibly some if not most operators
but for non-numeric inputs it may as well find the index instead
so hypothetically this could be complement under lowercase-alphabet (or bitwise-not under lowercase-alphabet if i go with 0-indexing) if the lowercase alphabet seems worth having as one byte
 
8:38 PM
0
A: Sandbox for Proposed Challenges

attImplement level-index addition code-golfrestricted-time Level-index is an alternative number representation to floating-point which claims to virtually eliminate overflow (and underflow, in its symmetric form). It expresses numbers as power towers: specifically, for any nonnegative number \$x\$, ...

 
 
2 hours later…
10:43 PM
@DLosc Thanks very much for the bounty!
 

« first day (4179 days earlier)      last day (666 days later) »