It occurred to me that an efficient-ish sieve isn't that hard with recursion. With this I was able to compute some truly gargantuan primes like 4999.

Primes←{0=≢i←⍸⍵:■⍸⍺⋄p←⊃i⋄(⍺∨p=⍳≢⍺)∇ ⍵>(≢⍵)⍴p=⍳p}_{(⍵⍴0)⍶⍶ 1<⍳⍵}

Huh. Line editor doesn't like tradfns with shy stranded returns, even though they're perfectly definable in the standard editor:

       ∇{(a b)}←f
 defn error

is there no dialect with a round function? I guess ⌊.5+ works fine too but it's just weird that nobody thought to add that as a primitive?

anyways I was thinking ⸠ for round in tinyapl, is it a nice glyph?

draft for tinyapl part 4 post

it's very code-heavy, it would probably benefit from some more explanations but they're hard :\

hmm, been trying to make more explanations but can't get them to be useful. anyways, any feedback is appreciated!

huh, chrome gives me a weird "link looks fake" error, i guess it's because the url starts with blog-rubenverg-com-...? anyways it's the correct site, it's the preview url deno deploy gives me

Uiua uses ⁅ for Round. I think it might have had some other related character in an early version though?

@Marshall as far as I can tell, it's always used ⁅, or at least it does since the first available tag on github

Could have been a long time before that (I have infinite Uiua hipster cred), but yeah probably not worth digging through the commit history for it.

realized I forgot to talk about the REPL! here's the second draft

the source code for the article is now over 1k lines long, parsing really was hard to do!

hmm, now that i think about it maybe i should've split parsing and interpreting into different articles? well, too late now, i guess.