« first day (3804 days earlier)      last day (1056 days later) » 

10:00 PM
Oh, it's that in-depth?
 
well just the domains that people are visiting from lol. not too precise
like for example, the plurality of people are referred to codegolf via codegolf.stackexchange.com (...)
 
I figured it was just summaries of things like question view counts, but also for things like the home page
 
codegolf.stackexchange.com/a/230711/95516 this is an interesting answer, is it definitely valid? beating golfscript in python...
 
It seems too simple... I'm trying to construct a counterexample.
 
@hyper-neutrino Technically speaking, 35k is the absolute max where after that there's nothing more to get with more rep
 
10:08 PM
Wait really?
What can you do at 35k that you can't at 25k?
 
Yeah. There's no new privileges after 25k, but you max out your delete votes/day at 35k
 
oh
well technically it takes 180k to max your flags, doesn't it?
 
Aw, Text doesn't have syntax highlighting for Fortran
 
Oh yeah, forgot about flags :/
 
@cairdcoinheringaahing You wouldn't be a very good Vyxal programmer then :p
 
10:11 PM
You get 10 flags/day by default, and then 1 additional flag/day per 2k rep to a max of 100, which you reach at 180k
 
Interestingly it does appear to support APL
And brainfuck?
 
The number of close votes/day is, for some reason, a flat rate
 
brainfuck isn't particularly hard to highlight :p though it's a bit weird they have that
 
Wait, there is fortran support
Lemme see what file extension it wants me to use
Oh hang on, Text sets the bindings separately from CodeMirror, so it might not actually support it despite having the code to do so
I might actually make a PR to add support for some of those
 
@rak1507 Actually I think this algorithm is correct. It's basically "try every color as the base color, recurse to find how many more steps each possibility would take, and use the possibility with the fewest steps."
 
10:15 PM
nice, very clever solution
 
In fact I think the 86-character Haskell answer may be using the same algorithm. (I don't quite speak Haskell, but it looks like it's doing most of the same things.)
 
I think a valid ngn/k port is {0|1+&/+/'o''x{x\y}\:x} which is pretty short
 
okay so I'm going to use a digraph adverb to extract functions to become first-class
between ɵ↑ and ɵ↓, which makes more sense to convert a function into a first-class object and which as an adverb to turn a first-class object representation into the function to call it?
 
First for first and second for second
 
okay, thanks - that's what i was thinking too
 
10:30 PM
ɵ is a neat character. Does it look right in monospace fonts?
 
it does
 
It's mine now. I'm stealing it.
:p
 
it kinda looks like ⊖
 
> applies adverb to builtin that i didn't implement
> adverb doesn't apply to that verb (since it doesn't exist) and uses the one before it (that exists)
*surprised pikachu face*
 
Oh wait, that's not an uppercase theta is it
It's circled minus
 
10:32 PM
mhm
i don't have uppercase theta in my codepage but i have the lowercase one
 
TBF I'm on 768p and 90% zoom on an 11" screen :p
 
my codepage uses small caps, greek letters, some armenian characters, some coptic characters, and some IPA characters lol
and it's incomplete for all of them because some don't fit or i don't like them or they look too similar to other ones
so i'll probably have digraph constants for the complete versions of those alphabets that i can't insert into my code itself due to them not belonging to the codepage
 
Ah, the old "I dropped all of the unicode on the floor and grabbed some random ones" codepage :p
 
pretty much :D
 
CMQ: Which custom codepage do you find most aesthetically pleasing, and why?
 
10:43 PM
@Neil I find Settings a bit more intuitive than Control Panel, although yeah, it'd be nice if they just changed Settings itself too
 
@DLosc In a lot of ways Husk's is nice, and sledgehammer's is nice due to the consistency
 
@DLosc I like Husk's. The characters are pretty distinct for the most part
Partially ninja'd :P
Actually, it's not really a ninja if we're voting
 
I like Risky's 4-bit one, but it's kind of hard to design a bad 4-bit codepage tbh :p
 
@RedwolfPrograms I bet I could :p
 
@DLosc I generally like Jelly's, mainly because the dots are distinct enough to stand out, and there's a dotted version of every letter aside from Q/q
 
10:45 PM
@DLosc Oh yeah? Well, I can design a bad 1-bit codepage! :P
 
@user :o
 
@cairdcoinheringaahing No there isn't
Not for all the dot types
(If that's what you meant)
Planning Ash's code page taught me that the hard way :p
 
No, I meant that there is at least one dot for every letter aside from Q/q
 
Oh, ok
Yeah it's annoying there's so few underdots
Although not really because underdots are ugly
It's the choice of ones with underdots that annoys me
is evil
 
Is this synesthesia based?
 
10:47 PM
Yes
Very ugly letter :p
 
1 is eviller, the evillest, in fact
 
I was considering making a language with a non-ascii compatible code page just to get rid of the es
 
@RedwolfPrograms Don't like the letter e? You might enjoy this novel.
 
@DLosc Plot twist: Redwolf wrote it after traveling back in time
 
if they went back in time to get rid of the letter e, they certainly didn't do a good job
 
10:53 PM
@lyxal I found this in a random folder on my computer earlier: i.stack.imgur.com/ijV1a.png
You may or may not like it
 
@totallyhuman Obviously, the forces of Satan are countering Redwolf
@RedwolfPrograms Turn it upside down to make it more relatable /s
 
husk's codepage is the best golf lang code page out there
 
In terms of aesthetics maybe
 
although in terms of non golf langs, APL's charset is nice, BQN's is also really cool
 
@rak1507 Why do you think so?
 
10:55 PM
But Jelly's is great
 
@user things like a product symbol being for product
 
@rak1507 Mmm, I don't like the similar-looking characters in APL
 
it's intuitive
 
@RedwolfPrograms The INTL keyboard thing makes it super easy to type
 
@Shaggy We still get semi-regular posts complaining about golf langs, as well as the odd newbie who wants to ban golf langs. As regulars we've established that golf langs belong here, but even now, there are some who only agree with that begrudgingly. The blog post is designed to present both sides of a long-running debate in the site's history, and fairly show arguments for and against. I doubt anyone is going to read the blog and walk away with the idea that we're open to banning golf langs
 
10:55 PM
@rak1507 Gotta agree there
 
whereas jelly's is just random and at times stupid
 
The only problem is with things like the trident thing standing for toadjX, but toadjX isn't very intuitive anyway :P
 
@rak1507 It's for a reason though
And a good one, too
 
@rak1507 Really? I've found Jelly's symbols to be the bit that makes the most sense
 
Jelly's code page isn't very mnemonic
 
10:56 PM
yeah, I get that, but I think how easy it is to type isn't a huge concern imo bc most people will copy paste anyway
 
@lyxal Time for you to advertise Vyxal's codepage here
 
@cairdcoinheringaahing maybe after using it for years
 
No, even when I started using it, it's mostly always made sense
 
@user Isn't that the one that had a homoglyph or whatever it's called for a while?
 
@RedwolfPrograms Oh yeah lol
 
10:57 PM
@user why? It's just Jelly merged with bits of husk, 05ab1e, Ash and other golfing languages
 
which is more sensible, ½ or √
 
There's lots of bits about Jelly to criticise as confusing, the code page is like the last on the list :P
 
@rak1507 ½
 
@lyxal Oh ok
 
@RedwolfPrograms contrarian!!
 
10:58 PM
@rak1507 For square root? Probably the second
 
Because it's easier to type, not as ugly, and still easy enough to understand
 
@rak1507 square root symbol
 
@user yeah, square root
first is jelly, second is husk
 
You people are animals
 
For circular roots, I prefer ○
@RedwolfPrograms Duh.
You're the only bot here
 
10:58 PM
½ is such an obviously better choice
 
Like I understand why ½ could be used, but the square root sign makes way more sense
 
@rak1507 That one is probably the most confusing, but even then, ½ makes some sense (sqrt = raise to the power of ½, ½ is typeable on USINTL, isn't etc.)
 
@cairdcoinheringaahing yeah maybe
 
Anyway gtg o/
 
10:59 PM
H vs ½
 
@rak1507 H kinda makes sense - Halve
 
P vs Π
@user yeah that one isn't so bad
I just think in general husk is nicer
 
Again, P=Product kinda makes sense
 
P is product
 
to me husk is a lot more readable than jelly
 
11:00 PM
@rak1507 Yeah
 
@rak1507 I use Π for product too
 
reminds me of my plans for neon
 
I've never found prefix languages to be readable tbh
 
@cairdcoinheringaahing Husk isn't quite prefix
 
@rak1507 I like the look of husk except for the characters that aren't the same width as ascii
 
11:00 PM
unfortunately going too far on the accurate symbols thing makes code look like a mess in most fonts
 
↑ and ↓ in husk are nice, considering they're the same as in APL
same with ← and → although they do different things (first and last)
 
@rak1507 Makes typing Husk a bit easier ←∪→
 
Jelly makes much more sense than 05AB1E or Japt
 
How so?
Are you talking about tacit vs stack-based vs "normal" or the codepage?
 
@cairdcoinheringaahing 05ab1e makes more sense than Jelly which makes more sense than japt which makes more sense than JavaScript.
 
11:02 PM
Code page
 
Oh
Code page wise I agree
 
Javascript's code page is the best, albeit limited
 
lol
 
Well, it's not a code page, but you know what I mean
 
05AB1E makes the most sense (stack = easy), then Jelly, then Japt to me if talking execution mode :P
 
11:03 PM
@cairdcoinheringaahing That's what I meant originally
 
Stack-based isn't always the easiest and kinda forces imperative solutions, but it's definitely easy to make and write basic answers with
For complicated stuff, Husk or APL are best (for me, I mean)
 
[rate my v1
 
I've generally found that stack based leads to hacky, "forced" solutions, whereas tacit/Husk-style prefix leads to more elegant versions
 
@lyxal On mobile, I see
 
11:04 PM
@user shush
 
Although, Husk often has to use weird workarounds for types/overloading that detract from it
 
@cairdcoinheringaahing Yeah, that's unfortunate, but the static typing is what lets it be parsed, so ¯\_(ツ)_/¯
 
@user laughs in experimental branch
 
@lyxal It doesn't look too bad at first glance, but the organization is a bit weird - superscript and subscript 1 and 2 are separate from each other
 
> ⎝⎠
ಠ_ಠ
 
11:06 PM
@lyxal laughs in terrible at Vyxal
 
@lyxal You've offset ASCII by 1
Space is 0x20, not !
 
@cairdcoinheringaahing Oh wow, very
 
Didn't Vyxal have two super big characters at one point?
 
@cairdcoinheringaahing that was fixed in v2.0.0a
@user yes
and its opposite
 
11:10 PM
Which version was that?
 
@lyxal Uh, that's very Japt-y versioning
going from 1.4(.x) to 2.0.0a
 
@Bubbler it was actually from 1.7 to 2.0 to 2.0.0a
 
Oh.
 
1.4 was just the version I thought could be the most cursed
 
Also known as the "Keg version" :P
 
11:13 PM
Because you know it's cursed when the release notes have the following:
> ſ now pushes 69 to the stack. That's because it's old function was absolutely useless
@cairdcoinheringaahing no, that's the K flag :p
 
@lyxal What an upgrade! 69 is merely nearly useless and not absolutely useless :P
 
@user that was when I gave a 50 rep bounty to Redwolf to add a 69 constant to Ash
Heh, what a waste of a bounty
 
Redwolf played you there :P
 
lyxal: I'll give you 50 rep to do this thing to your language
Redwolf: Does thing to language, decides it's so garbage now that he just scraps it
3
Any feedback? (Codidact sandbox)
 

« first day (3804 days earlier)      last day (1056 days later) »