« first day (3816 days earlier)      last day (1018 days later) » 
00:00 - 23:0023:00 - 00:00

12:04 AM
I do like that the most starred message in the Jelly room is
in Jelly, Apr 1 '16 at 18:42, by Dennis
I don't understand my language. ._.
Which is in context to LCCs (and dyadic links) :P
 
12:40 AM
*sigh* yet another issue with the wiki
> Ƭ Like ÐL (alias for ÐĿ). Collects all intermediate results.
fun fact: Ƭ does not behave like ÐĿ
 
12:57 AM
In fact, Ƭ<tail> can be just as useful as ÐL when using dyadic chains
 
well, this finally explains why i've never understood like half the quicks
TBH until recently i didn't realize most of the loop quicks varied the right argument per iteration until i looked through the source code / you mentioned the fibonacci behavior
 
I understood most of them either through experimentation or by asking Dennis tbh
 
and explains why sometimes ÐL doesn't work but ƬṪ does lmao
making this resource has been an excellent way to understand the language myself, TBH :p
 
What's the saying? "The best way to learn is by teaching"?
 
something like that :p
aight, added all the quicks to the data table too :D now i just need a few syntax components and i should be done the boring part
given how easily I give up on things, I'm surprised I managed to enter nearly 4000 lines of JSON by hand (a good amount of lines just containing { or } tho)
 
1:10 AM
5 hours ago, by N3buchadnezzar
The best way to learn is to teach
 
1:35 AM
are there any loop quicks that don't fibonacci other than Ƭ
 
1:47 AM
from what i can tell, no
 
 
2 hours later…
3:39 AM
CMQ: what is a good favicon for a jelly-related site?
 
0
A: Sandbox for Proposed Challenges

AviFSCompile Esolang to Esolang Based on Convert FRACTRAN to Brainfuck. Compile Brainfuck to FRACTRAN Compile /// to Brainfuck Compile Brainfuck to /// If you have other ideas, please add them. Be sure it's reasonably doable with simple tarpits, eg Jelly to Vyxal to APL isn't all that fun. If you ca...

 
@hyper-neutrino Is there any operator particularly associated with Jelly?
Maybe one of the dotted letters
 
not IMO, but i'm thinking about that
unfortunately there are no dotted J characters
 
I thought there was an underdot one
Oh wait no
There's an "overdot" one where the dot is removed instead
Maybe a J with a dot is a good idea though, since that's not a real operator
 
yeah i guess ȷ could work
though that just looks like a normal capitalized J with no context lmao
 
3:52 AM
I was thinking something like J̇ (if that renders right), despite it not being a real character
 
@hyper-neutrino just have jelly as the picture
 
@RedwolfPrograms wait what lol. interesting
i could do like clipart jelly
in which case, what color?
 
Red seems like a good choice
 
I think Red is pretty much the typical Jelly Colour.
 
(for reference, the entire site is light blue, but it doesn't have to match / arguably shouldn't)
red seems like a good choice to me too
 
4:01 AM
Red also could plausibly be either type of jelly, unlike light blue
 
Try to use a Shade of Red that compliments the blue.
 
might have to go a bit in the pastel pink direction for that
 
A 120° shift might look good, haven't checked though
Might end up too purple or orange
 
also another icon idea
ƙ
 
i can't find any satisfactory ones on google images so i might need to go to another image search site or just make one
@UnrelatedString aaaaaaaaaaaaaaaaaaaaaaaaaa
 
4:03 AM
it's visually distinctive in that i don't think other codepages tend to have those tailed letters
and then it also has
cultural significance
 
0
Q: create an array using for loop and user validations

Bhadar ChaudharyI have tried to solve this problem but cant seem to figure it out please help. the program using an already given array and one user input array and for the third array the program takes the user input and inserts that into the power formula the result is displayed all in ascending order in table...

 
It just has to be a 16x16 pixel art for a favicon, so not too difficult (especially if all you need to make is a blob in a bowl :p)
A 120° shift looks more pink than red it seems
 
@NewPosts probably would be instaclosed on SO so i will just follow the Don't Migrate Crap rule and let it die
 
also looks like hw lmao
 
4:07 AM
better suited to so does not mean suited to so
it is 105% homework
 
I spent wayyyy too long on those links, but I just converted my comment to an answer for the FRACTRAN self-interpreter question!
@cairdcoinheringaahing :)
@NewPosts Uh oh!
@hyper-neutrino Ooh, I made one once
I used a free jelly/honey font
 
How do people keep finding Code Golf and assume that we'll solve their problems?
 
Sort of drippy and gloppy looking one
 
@ATaco idk we even added a super detailed first-time-asked dialog to try to stop people from not getting the point
you can only make so many resources; people have to actually read xD
 
4:20 AM
I did something like this, with a light brown background
 
Wouldn't look too good as a 16 pixel wide favicon though :p
 
It was for this, but I clearly never committed
I don't know anything about favicon, can you tell me?
They're like the TIO in tio.run's corner, right?
Or Dyalog's DYALOG in alien sci-fi font
 
Not sure what HN's looking for specifically but usually I hear favicon in reference to the icon by the title of a tab in the browser
 
Yeah, but it sounds like maybe you know stuff about their design that I don't
Eg. I didn't associate them with being low-res
The TIO & DYALOG don't look 16px to me
But I might not be understanding
 
This is what I think of when I hear favicon
 
4:24 AM
I might be getting it confused with logo
OOOOHHHHHH
I always thought it was this:
Oh wait, that is also the favicon for me, but the DYALOG favicon is def way simpler than their logo
 
uh yeah i mean the icon in the tab
 
I think I'm using the words right, now...
And those are restricted to 16px?
 
they're def not restricted to 16x16 are they?
that circle for TIO looks way too circular to be a 16x16
 
jinx
 
i've played minecraft, i'd know :p
 
4:26 AM
Hahaha
 
Not sure
 
Tat wasn't the favicon, though
 
mine are usually 256x256 i think. but they definitely get downsized
 
This is, lmao
But still not 16
 
TIO's is 32×32 it seems
256 is definitely overkill though, I doubt there are any screens where the favicon takes up more than 32 or maybe 48 pixels
 
4:28 AM
> Size: The optimal size for creating a favicon is 16x16 pixels, which is the size in which they are most commonly displayed. However, they can sometimes appear in larger dimensions too (such as 32x32 pixels). If you need help getting the right size for your image, try using Wix's Image Resizer tool.
 
256 is definitely too big; i think i sometimes downsize them to 64 or 32
 
I just rickrolled someone over walkie talkie
My life is complete
 
lmao, nice! do tell
How about the quotation marks for jelly's favicon?
“»
It's kinda cool, maybe...
 
that could work, since i think that form of string compression is unique
 
@AviFS a friend of mine's kids (this is an older friend obviously - think older than Dennis) went out bike riding. They had to take a walkie talkie with them for communication (the kid is around 10ish), so I got a hold of the walkie talkie on the end here and epicly pranked them.
2
However I suspect my efforts may have been wasted due to the victim not understanding what a rickroll is
 
4:33 AM
How about something like:
 
hmm. not sure - on one hand, (other than the fact that i ripped off most of jelly's structure and just extended a crap ton of features for yuno) this is unique, but on the other hand, it isn't that representative at face value
 
@hyper-neutrino how about the compressed string for jelly?
 
“Ẏṃ»
 
Yeah
Like that
 
idk if i can fit "JHT" into a favicon - this site is unofficially (tho i could probably get support for it) a JHT site
 
4:38 AM
I have a lot of different fonts I generated, but they're all a little too cuneiform-heiroglyphic-Chinese/Japanese/Korean-looking
I guess I'll just post them, though. Here's a page. Incoming
 
huh, interesting
 
ah, cool (i'm gonna collapse your images to save space)
 
Is that possible?
Oh, thanks!
Well, but then no one will open them
They're less likely to be seen
 
right but 5 images takes up too much space :p
 
4:40 AM
But anyway
 
that's cool tho
 
It's obv is't going to be confused with all those other ancient-and-less-ancient scripts, haha
 
i think the third is prob my favorite of these
 
But the point is I think that general idea might be slightly too "exotic" "foreign" "not-easily-recognizable" looking?
It does have a bit of a scratchings in the sand/carving in stone/calligraphy on handmade paper feel which you may or may not want
 
i'm not sure how i feel about that part - idk if it would be representative/recognizable tho
 
4:44 AM
@hyper-neutrino Yeah, that one is nice because they merge together a bit. The >> is almost at the same height as ""
 
also with the 4th+5th one the quotes look like they're backward :p
 
But some of the others have more motion, which I like
Some look like they're zooming
 
hm, good point
 
Particularly 2 & 4
Even 1 has a lot more motion than 3
@hyper-neutrino Which part?
 
4:45 AM
like the sand drawing / carving look&feel
these kinda remind me of tsh's pfp style
 
Oh, that's the feel I get. Or like some sophisticated calligraphy
 
yeah
not sure it fits with the (attempted) modern site style tho :p but it does look quite nice IMO!
 
@hyper-neutrino Were you specifically objecting to me saying it reminded me of a Mesopatamian or East-Asian script? Or just that you don't have the same general feel?
@hyper-neutrino Haha, thanks. Glad you like it! Was totally out of left-field, tbh
I hit Hello World on TIO to get a sample program that I could base it off
And, of course, the quotation marks were the only interesting feature of such a simple program
 
@AviFS oh, no, i agree with that - i was saying i'm not sure how i feel about having it for a favicon :P
 
Oh, yeah me neither!
I don't really like it, haha
That you like it at all is nice to hear, haha
I kind of like the idea of jelly clipart, tbh
I think this is the most doable kind of jelly to 16x16ify
 
4:58 AM
yeah, probably
 
Though I'd call that Jello. (California, US)
@hyper-neutrino Hadn't even sent it :p
I mean jello is more doable than a jar of jam
 
I think Dennis was originally going to name the languge Jello but changed it due to potential copyright or smth
@AviFS oh oops i just assumed you meant that :p
 
Wait, really??!?!
 
idk. don't quite remember
 
@hyper-neutrino Oh, true, you're right. That was most of them in the google results
Alright, gtg but be back later
 
 
1 hour later…
6:12 AM
0
Q: Generate this huge right triangle

js2022Your task is to print out this exact text: - || -|- |--- |-|-- -|-|-- -|----- |-|----- |---|-|-- -|-----|-- ---|-|----- |---|-|----- |---|-----|-- -----|---|-|-- -|-|---|------- ------|---|----- |-|---------|-|-- ---|-----|---|---- -|-----|-|--------- |-|---|-|----------- |-----------|---|-|-- -|...

 
7:08 AM
hi everyone
 
7:47 AM
Does 05AB1E have a mold builtin??
 
7:59 AM
i don't remember such a thing
 
Bad then
 
like, which type of mold
mold like another array?
 
@Razetime probably this
 
yeah i doubt it
Jelly's mold is sorta reminiscent of J and APL reshapes
most other golfing languages wouldn't have it
 
Laughs in molding in Vyxal being broken with generators
 
8:06 AM
@Razetime I thought it would fill an array, replacing its data but keeping its structure.
 
and reuses contents if necessary
 
Right, so it is more like ⍢∊ in APL.
 
under enlist?
hm i guess
 
vyxal's mold is horribly cursed
 
that symbol has a lot of overloads
 
8:09 AM
what does a mold do?
 
there have been a lot of times when i've wanted to use a command but it does something else instead of vectorizing bc of overloading
 
@PyGamer0 Fill an array, replacing its data but keeping its structure.
 
prob just personal preference but i do not like overloading of things that would be useful to vectorize, which vyxal has a lot of
 
you do have a point
i'm sure a lot of trimming can be done
i dislike some husk overloads due to type inference spazzing out
 
@hyper-neutrino the function or the command?
 
8:11 AM
@lyxal idk all i know is the built-in never does what i want
 
@hyper-neutrino well I stole the mold function straight from jelly
 
vyxal range built-ins should auto-vectorize btw
instead of just dying
@lyxal then sounds like generators are just cursed. oh wait that's well known xd
 
@hyper-neutrino open an issue on the repo
Tag it appropriately
@hyper-neutrino there's a reason the tag description is lyxal: *cries*
 
Generators are the polar opposite of Rick Astley: they give me up, let me down, make me cry and hurt me
10
 
8:20 AM
i ended up getting 40ʀ¦›2lƛ÷ræƛkA$i;ṅ, for the triangle challenge (+2 over wasif's)
well, 1 if i use j
there are like 5 built-ins there that don't do what i want that could've taken much fewer bytes lmao
 
Which ones?
 
i not vectorizing is the main one
tho if mold worked this could be shorter
more so i'm just not used to this builtin set cuz you don't have pairwise
 
I should make l do pairs if it's first argument isn't a number
 
yeah that'd make sense
 
8:32 AM
@hyper-neutrino wait wdym by that?
 
notice how i ƛkA$i; instead of just kA$i
 
Does İ work?
 
oh wait it does lol
still doesn't actually save over wasif's (it's 1 byte shorter but i can suggest that as a golf i think, even though i have different sol)
anyway it's 4:37 i'm gonna go to sleep now o/
 
Good
 
8:53 AM
Back again
@hyper-neutrino o/
@PyGamer0 why the rewrite??
@Razetime yes that's what I meant
 
 
2 hours later…
10:48 AM
@wasif i have stopped the rewrite (it was for organising the bad code)
 
I see
But why did you stop at only 97 builtins
 
@wasif i would add more
 
nice
 
11:08 AM
Anyone have a good post I can put a +100 bounty on to get my rep down to the perfect amount?
 
@pxeger maybe this
 
⁶ Fourth Input / Sixth Argument Atom niladic (0) Get the first input (third command line argument), or " ".
 
???
 
11:49 AM
@hyper-neutrino Yes, he was but couldn't because of the copyright :(
@Neil Is that from hyper's Jelly docs?
 
12:05 PM
Dilemma: I'm trying to create a language that truly minimises redundancy by making it basically impossible to create two programs that do the same thing. However, I strongly suspect this will take all the fun out of golfing if there is truly only one way to do something because it will be impossible to ever improve a program
 
12:20 PM
0
Q: Will the hydra finally die? Part II

N3buchadnezzarBackground This is a follow up question to the question: Will the Hydra finally die? As before a dangerous A Medusa have released a dangerous Hydra which is revived unless the exact number of heads it have is removed. The knights can remove a certain number of heads with each type of attack, and ...

 
@NewPosts 👀
 
12:56 PM
@pxeger I suspect that's impossible
 
yes in a TC language it's probably not really possible but we can get to some extent asymptotically close
for some definition of close
 
If you have any two builtins a, b which are inverses of each other, then an arbitrary program can include ab and do the same thing
And if you don't have such builtins, how do you intend to increment/decrement?
 
I syntactically prevent ab from existing (not make it an error, but make it unachievable)
how exactly I do that is a trade secret for now lol
Because increment/decrement is more common than adding or subtracting 53, I have builtins for those
But as a result I make it syntactically impossible to do 1+, and that "entropy", as it were, is taken up by some other behaviour
 
@cairdcoinheringaahing yeah there was also a string missing its closing quote somewhere
 
1:59 PM
@pxeger i've had that idea too and it's definitely interesting to think about
i think strictly speaking it's impossible because if you have code that solves a problem in one domain it also solves it in a smaller domain but it would also be solved in that smaller domain by something which perhaps even hardcodes different behavior outside that domain
and there's probably some other problems if you sidestep that consideration
 
2:27 PM
@Neil fixed both issues, thanks
 
 
2 hours later…
4:17 PM
@pxeger ngl, this sounds kinda similar to an "inverse Stack Cats", in that in Stack Cats, every builtin tries to undo itself, but in this, no builtin can undo any other builtin
 
stack cats definitely comes to mind yeah
except this is more a case where you have builtins that can undo each other but it's impossible to use them to do so
i'm assuming the encoding scheme is akin to some kind of contextual frequency compression where the probability so to speak of redundant behavior is defined as 0
 
4:56 PM
more or less, yes
 
5:20 PM
Why are people from Norway so good at editing files in Linux?
Their ancestors are vi-kings
 
5:38 PM
0
Q: Most Valuable Circle Chords

JonahThis challenge was inspired by this tweet. Idea Consider a circle with n evenly spaced dots around the perimeter, where each dot has a postive integer value: Now connect the dots with non-intersecting chords: Next consider our score, which is the sum of the products of these pairs. When n is o...

 
5:51 PM
@hyper-neutrino Are you considering extending your integer / list compressor to include the builtin constants (e.g. it says the shortest way to represent 1,2 is 1,2 but ؽ works)?
 
@cairdcoinheringaahing eventually, I intend to make it also be able to determine ways to use grade/group, base conversion, etc, but that's a project for the long term
but yes, built-in constants is an earlier step
the main reason I am restricting to literals is a) easy to implement and b) the user can tack on other literal components easily, but that latter part isn't all that necessary if the user just uses a couple of ¤s
 
Tempted to write a challenge about Jelly literal meta golf :P
 
also, i might try to merge my integer and string compressor to just make a "constant compression tool" but it'll be a large project if i want to make it even close to fully optimal
if lynn's brain had an API this would be easy :d
 
Man, brain APIs would be such a cool (and weird) concept :P
 
@NewPosts i feel like there's some approach with like get all partitions for each rotation and recursively apply until you've broken it down but that also feels a) slow and b) long and i'm sure there's a more elegant way to do it so i will hold off on my fgitw attempt
 
6:07 PM
Maybe I shouldn't advertise my own answers but I'm stupidly pleased with codegolf.stackexchange.com/a/231230/3852
 
hyper was just remarking on that in jht
 
How often do you get to golf down 40R ? That feels like such a once-in-a-lifetime trick B)
oh I should like, go be in there
 
hm, on second thought i probably won't try to make a more optimized nilad compressor even in the future
a) sounds like extreme pain
b) if a computer can do all these big brain golfs then they stop being as impressive :c
 
@hyper-neutrino No pain, no gain amirite
@hyper-neutrino Making such a computer would be pretty impressive
 
very true but i'd rather spend that time gitting gud at jelly :p
 
6:21 PM
cmp; git gudding, gitting gud, gitting gudding, or gudding git/
 
doing the git gud
 
Forgot that one
also mercurialing gud, mercurial gudding, or doing the mercurial good
 
@hyper-neutrino I think adding constants is almost a must tbh
 
constants, yes
and probably a base compression util too
but i don't think i'll go so far as like group/grade, except maybe some basic things like if it detects the list is a permutation or a grouping of indices, then show the ungraded form / permutation order and tell the user to look for a smarter approach there
 
@hyper-neutrino git config --global alias.gud '!echo lol'
 
6:34 PM
I like answers which are like "this would run if you had 40T of ram available"
 
"x"*2**45?
 
@hyper-neutrino i tried this in javascript, it ends up to be shamefully long
 
@hyper-neutrino I might have a look at some constants which are "nilad applied to monad", and compile a list for you
For example, Jonathan's trick with group to produce [[1]] in either a niladic chain, or when just passed an integer argument
 
@cairdcoinheringaahing that'd be helpful, thanks
i could even add like an index where verified people can add a constant / constant category they find based on some pattern, but idk how well that'd work
i'll look through the atoms and try to find a list of things that are common enough and not insanely hard to reverse-engineer optimally
 
7:14 PM
@user gud-gitting
 
 
1 hour later…
8:39 PM
0
Q: Classify a graph

hyper-neutrinoChallenge Given a graph (a structure consisting of nodes and vertices), classify it according to a few categories. Specifically, you will be given an unweighted directed graph, which is a set of nodes and edges from one node to another (but not necessarily both ways). You can take this in any rea...

 
@NewPosts can this be a ?
output is not binary but the tag just says "For challenges which involve deciding whether or not the input meets certain criteria, and outputting some data, representing that decision.", and this has multiple decisions with the output based on those
 
i think i do remember precedent for that interpretation
 
Upvoted!
> If the graph is undirected, this checks if the graph is a tree.
> If the graph is undirected, this checks iff the graph is a tree.
 
disagree
 
Too small an edit for me to suggest at <2000
@hyper-neutrino by def
 
8:44 PM
iff makes sense iff the left and right side are equivalent boolean expressions IMO
"check if <condition>" is an english expression
 
> In graph theory, a tree is an undirected graph in which any two vertices are connected by exactly one path, or equivalently a connected acyclic undirected graph
@hyper-neutrino Oh wait, it's missing the connected part
nvm
 
9:03 PM
@xnor Regarding my previous comment about OEIS submission: I just found out I can create the entry with my formula, and then you can edit it to add yours. So it will be automatically attributed to you. Do you have an OEIS account?
("found out" meaning @flawr told me it can be done)
 
9:15 PM
I really like the last few codegolfs that has popped up =)
Anyone watching some foobal? :P
 
Not since Spain was eliminated :-)
 
No because I don't think anything good will happen here, regardless of the results
Either there will be drunken riots in celebration, or drunken riots in anger
 
Tbf it seems you guys drunken riots over everything
2 TABs good! 4 spaces bad!
 
1.5 tabs, 0.5 spaces, duh
 
1 space or bust
 
9:20 PM
No indent, embrace chaos
 
s/chaos/C
 
Use backspaces and vertical tabs as indentation
 
I made a strawpoll for TNBers to vote on which is best (multiple choice). Once a decision is reached, let us ban all other indents!
 
This might lead to...
another inci dent
 
I've also disabled duplication checking, so go crazy >:)
@N3buchadnezzar Another incident?
 
9:31 PM
What is this first time user modal I keep hearing great things about? 👀
 
open a new private/incognito window and click "ask question" to find out :p
 
9:54 PM
Btw, whoever managed to click 1 spaces 6969 times, how did you do it?
I tried writing this, but it just reloads the page continuously, clicks at the end, and stops:
 
@user sent the web request
 
function food(e){
  window.location.href="https://www.strawpoll.me/45481790";document.addEventListener("DOMContentLoaded", e=>{
    document.getElementById("field-options-10-tabs-and-2-484-spaces").checked=true;
    $("button:submit")[0].click();
    food(e)
  })
}
@rak1507 Oh cool, how do you do that?
 
sent it, view network, copy as fetch, then just use a for loop
 
Ah, that's smart
 
@rak1507 Do you have the code?
 
9:58 PM
if you vote + open the network tab, then you see a post request, right click + copy as fetch
then paste that in the console
 
10:30 PM
lol 22.5k for 10 tabs
 
and 2.484 spaces
 
yeah
 
Trying to reach 42069
 
10:40 PM
lol how did No indentation whatsoever reach 50k+?
 
because i ran a 0-ms setInterval 8 times in my browser
 
My tab nearly crashes if I send more than 1500 requests at once
 
10:52 PM
TNB has spoken - 10 tabs and 2.484 spaces are the best (I'm discounting HN's 58,572 votes for "No indentation whatsoever" because they ruined the beautiful 6969 votes for 1 space :/)
 
:(
how could you
 
It's time to make "No indentation whatsoever" 69,420 to right this wrong
 
00:00 - 23:0023:00 - 00:00

« first day (3816 days earlier)      last day (1018 days later) »