« first day (1606 days earlier)      last day (22 days later) » 

5:36 AM
@RubenVerg so what I was going for with the funny triangles and dots was a visual metaphor of which side has the drop on
Meaning as long as the dots are kept on the side they're on, it's fine
@RubenVerg honestly if it looked like :> as a single character that'd be good
Because it's intended to be the same as x[1:] but slightly visually different to indicate the behaviour is slightly different
Also, I've had an idea
The glyphs you're making will be the default font on the website
But I might make a variant font where each symbol is represented by what I wish I could draw if Unicode were something I could forcibly add to
So like each glyph would be a better representation than what is available in Unicode
The point being that when you want to edit things in a way that looks like what you'd get in other fonts, use vyxal386 (be great if we called it vyxal420 though)
But if you want an easy mode, use the easy mode font (call it vyxal69)
It'd be good to have both fonts. One for preview of what other people see, and one for beginners
Professionals use the 420 variant, people getting used to things would use the 69 variant. 420 is something you'd put in a museum, 69 is what you'd whip up on a napkin while eating soup at a petrol station diner.
I am a responsible font namer
 
4 hours later…
10:04 AM
@lyxal currently just modifying tinyapl386, not making a specific vyxal386 font
do you want a special font without all the extra stuff from APL and BQN and TinyAPL?
i think while i'm developing it i'll keep adding stuff to tinyapl386 and then if you want it to be split i'll make a cut down version at the end
10:22 AM
@RubenVerg what's this then?
@RubenVerg ah got it
I mean, if it's already a font, splitting it isn't hard I imagine
10:42 AM
it's not, but i think mantaining two projects is unnecessary
unless you want something that's a breaking change from tinyapl386
also this question again mostly because i don't think there's a nice way to distinguish them anyways in a monospace font
I was saying it's more a shame that so many Unicode chars aren't single width
We're well beyond the point of hoping to have everything the same width :p
@RubenVerg I'd rather not be lumped in with APL :p
I could always just do the splitting myself or via workflow
11:08 AM
@lyxal alright(:
@lyxal yeah that's probably a better idea
everytime i have to maintain two copies of something i inevitably make changes to one but not the other
12:06 PM
vyxal font??
theseus will have support for this
12:45 PM
some more (most notably )
hey so is that intentional to have 2 sets of underlined 1, 2, 3 and 4?
you do in the codepage (: is used both for the grouper and for the constants
ah
very inconvenient
i assumed constants were temporary so i didn't tell you
they are not (I can not be bothered to find that many appropriate symbols :p)
12:49 PM
the :V brothers
quirky little guys
@Ginger there's also a cousin down in e0
@lyxal maybe you can use the double circled ones (i'd probably make them double underlined)
webpack time
@RubenVerg fixed
I love import monkeypatching!!
12:56 PM
oh, i didn't notice æ for :⇄, it's really nice!
the idea being it's roughly mirrored :p
yeah i can see that
i wanted to ask, "does it do :⇄ or :dip⇄" but i can't find the symbol for dip
still not done or just renamed?
forgotten :p
off to the symbol table I go
bouta commit a crime
⍤ for something that isn't function composition
>changes webpack buildscript
>compiles without errors first try
wha
but did it monkeypatch the plugins correctly
it did not
1:11 PM
@lyxal this is a smaller crime than for something unrelated to complex numbers
oop
I think it worked??
it seems to have worked!
@lyxal good to mess with the workflows?
sure
oh joy!
there, done the :v brothers
what's the difference between E and ?
Eval and Exec
1:23 PM
yeah. what's different between the two of them? is it the same difference as python?
it worked
@RubenVerg correct
except exec here is vyxal exec
eval is a hand-written eval function
that was a lot more anticlimactic than I expected it to be :p
1:25 PM
the action worked second try
hmm, still no way to input syntax stuff?
no ctrl+enter to run?
why are you telling me these things now :p
because I didn't test it before :p
@lyxal shift-enter is run
I think there was some issue with ctrl-enter
1:27 PM
@Ginger i think i mentioned inputting syntax glyphs at least twice before
I'll add them to autocomplete, one second
does pressing tab not autocomplete?
no, that's default CM behavior for accessibility reasons
it's Ctrl-Space
how do I cancel a suggestion?
@RubenVerg added
@lyxal wdym?
1:37 PM
if I'm typing a lowercase letter, the editor loves to make it uppercase after I press enter
but sometimes, I do not want it to uppercase it
escape closes the suggestion list, but it uh
shouldn't be doing that
I forgot to turn on case sensitivity in the searchers :p
fixed
also, no auto-run?
notice how the program runs upon opening the link
ohh
yeah I can add that
just gotta figure out where to trigger it lol
would be Super Cool if use() were stable
oh dear. I broke something
1:52 PM
Woah v3 is sort of cool.
More cool than anything I could ever make.
@Ginger it is not?
I have a vague memory of seeing it in websites.
OMG.
This is even better now!
the, uh, dictionary is not being fetched correctly
because I goofed up the config
Wait.
Well, wouldn't it be better if you could pass infinite arguments to +?
Take 4 5 8 + would be 17.
And 4 5 8 + 2 - is 15 of course.
how would you determine when to stop?
Whenever we get the + sign?
what if I only want to add 2 numbers?
1:56 PM
me when I use DATA_URI for both the dictionaries and theseus.json but forget to stop hardcoding it to just theseus.json
Only do 4 5 + then.
@192927376337929292283737373773 but I might need to have a third number
what if I want 10 - 4 + 5?
10 4 5 + - is going to take too many things
Honestly I don't get the logic.
Give me one minute.
1:58 PM
dictionaries are fix
Because that's how things already work.
You just need to instead of making it a dyad, make it variable length.
I'm saying that with + taking infinite things, it no longer knows to not take the 10
I don't want it to take the 10
You can rephrase as 4 5 + 10 -.
That's -1 though.
Hmm.
correct
This is why bracket syntax is more used.
There you can simply do -(+(4, 5), 10).
2:03 PM
well that defeats the point of a stack language doesn't it.
But Vyxal is a golfing language.
And stack.
Well if I want to add three numbers what do I do?
++
amazing, I know
...
Why doesn't that have documentation?
...because it's two plus signs
+ +
add two numbers, add a third to the result of adding two numbers
But 10 2 3 ++ works for me.
2:07 PM
yes
+ + == ++
spaces don't mean anything
I have programmed myself into a corner
Alright.
But that should be documented in the Addition thingy at least.
Right now it doesn't show that special syntax.
it's basic stack mechanics
it's not special syntax
Oh.
So we're pushing 10, 2 and 3.
And the first + adds 2 and 3.
And the second + adds 5 with 10.
OHH.
ok what if: instead of making the code to execute a parameter to VyTerminal's imperative handle's start(), I make it a prop and have start() take no parameters
which means I can run code if it isn't empty when VyTerminal mounts
also VyRunner should probably handle timing out itself
2:13 PM
I really hate Replit man.
hey ginger could there be a link to github.com/Vyxal/Vyxal/blob/version-3/documentation/table.md on the 📓 tab?
sure, where?
maybe
while there's on-site search, I still find myself going to the table a lot
anyhow, in terms of codepage formation, I've created a text file to start figuring out which functions still need to be added, and which functions are just overloads
doesn't contain any new info at this stage
it also doesn't contain syntax things
I've info'd up to A, with more to come tomorrow
I think I'll be off for the night now
o/
@lyxal done
I am now intending to build a new language.
+(2, 2) = 4
-(+(2, 2), 1) = 3
"Hello, world!" = Hello, world!
Btw brackets are optional.
So +2, 2 and -+2, 2, 1 work.
But uh, brackets are highly recommended.
2:32 PM
congratulations, you've invented Polish notation :p
Thank you. Thank you.
I am not a Pole btw.
okay, version archive workflow changing time
this is the hard bit
3:05 PM
@Ginger hmm, how do I input ƛ?
oh, if i search for lambda it appears but if i search for map it doesn't
I think the archive works??
oop
it almost works, but I messed up the paths in the webpack config
3:38 PM
it works!!!
 
1 hour later…
4:48 PM
more progress
btw if you want to check out the font you can download the latest here
once the codepage is settled i'll add it to the showcase as well
for now they're secret extra glyphs(:
 
1 hour later…
6:16 PM
version list and header links have been added
7:14 PM
3.5 is gonna be great
 
3 hours later…
10:00 PM
Random thought: find / truthy indices (/ index?) could totally be multidimensional by default
I don't think I've ever used i's slice overloads
10:50 PM
@emanresuA find, for sure, truthy indices should probably stay one dimensional, and index I think would be a good idea (slice could be a digraph)
Actually maybe not find
But index I can see being a good idea
I can't think of a case where I've needed to take the indices of nomempty sublists of a list
And for find it'd just be a natural case of depth switching - any other interpretation wouldn't make sense
11:40 PM
depth switching for find is good
I don't know whether making it multi-dimensional is good though
11:51 PM
What else am I going to do with finding an int in a 2d array of ints
returning -1 for not found seems like something that's useful in formulas sometimes
I'm open to adding it, I need a bit of thinking time to be convinced that it won't ruin some useful edge cases

« first day (1606 days earlier)      last day (22 days later) »