@AviFS I like to think that we have very high quality on our problems, that is certainly not always the case. And, we support APL, of course :)

I think your output testing code is way too lenient - if the output should be 1\n2\n3, 1\n 2\n\n 3 shouldn't be allowed

cfr chat from here

and specifically the 25 (utf8) byte sol {⎕←⍵/⍞}¨2|⍳⎕ which prints empty lines for even lines

@RubenVerg Yes, the output testing is very lenient on "whitespace change". This is on purpose though, I think it's a good thing.

yeah, idk, I guess on one hand it makes golf more annoying to beat others of the same language (takes a while to realize one completely different approach has compatible, but not exact, output), but on the other hand the emphasis on code length doesn't seem that important

unrelated, but did the monadic Base as described in A Dictionary of APL (tokenize the input) ever appear in any dialect (I'd assume so, the Dictionary describes itself as a collection of things dialects did at the time), and why did it then disappear?

Yes, exactly. We are not primarily a code golf site. We focus more on correctness first, and efficiency (runtime) second, and less on code length.

But isn't being lenient on whitespace the opposite of focusing on correctness, at least in problems where whitespace plays a large part like open.kattis.com/problems/echoechoecho?

Btw, you can't see other people's solutions, right?

Anyway, I'm really enjoying the problems. Thanks for the site :)

@RubenVerg Not as ⊥, but J has ;: for that and Dyalog (unofficially) has 60⌶

@Adám 60⌶ would have been useful to know. Is it undocumented for a reason?

How many more I-beams are undocumented?

@PaulMansour A quick count gives me 64 fully documented I-beams, 12 undocumented ones are included on APLcart, and a further 178 are entirely undocumented.

@PaulMansour Not really. It could possibly change behaviour, but then again, so can any I-beam.

@yairbiton Hi there. Interested in APL?

Good to know. I'll guess I'll go research the 12 "undocumented" ones on APL cart.

@Adám doesn't that make them documented then... :)

That's why I put it in quotes.

The next question is how many of the 178 should be documented, versus how many are really for internal use only.

@PaulMansour Here you go.

@rabbitgrowth the 14 byte sol here is an abuse, just in case it wasn't clear

@Adám this doesn't seem to do the same thing (specifically, stranded arrays are in the same string in Dictionary's ⊥ and different strings in 60⌶) but still really useful, thanks

@Adám if I guess what an i-beam does, will you tell me its number?