Yeah, such a shame. I feel like someone should make a mini documentary or do interviews with original APLers, I think it'd be really interesting and good to preserve some history
Seen that one too. I wonder if a shorter (5-10 minutes) video similar to one of these but more compacted would draw some attention to APL like the game of life video and similar
<moon-child> looking at that video, I wonder: might the tradition of golfing originate, at least in part, out of a desire to conserve ink and paper when using teletypes?
I was toying around with a recursive prime sieve and I wonder why this gives me a WS full:
s2←{22≥⍵:⍵(≥{⍺/⍵}⊢)2 3 5 7 11 13 17 19 ⋄ p←∇ r←⌊⍵*0.5 ⋄ c←1+r+⍳⍵-r ⋄ c/⍨∧⌿0≠p∘.|c} for s 1000000
but s ← {22≥⍵: ⍵(≥{⍺/⍵}⊢)2 3 5 7 11 13 17 19 ⋄ p ← ∇r←⌊⍵*.5 ⋄ c ← 1+r+⍳⍵-r ⋄ c/⍨⊃∧/p×⍤|¨⊂c} doesn't.
(⎕IO←0)
The only difference between s and s2 is the final "statement"
It seems to me that p∘.|c and p×⍤|¨⊂c generate pretty much the same data, except the outer product creates a matrix and the ×⍤|¨ creates a nested vector.
I also rewrote the last step as c⊣{c/⍨←×⍵|c}¨p in the hopes that constantly reducing c for each known prime would not make it blow, but this version doesn't go much farther than the ×⍤| one
@EliasMårtenson I don't know what one half bit means. Are you not considering even numbers, is that it? And I am interested in an implementation, not on the idea.
@Adám True. :-) I used millibits, because it's a pet peeve of mine when people write mb when they mean MB or Mb. So the previous example would be the first time I actually used the term mb in conversation :-)
@rak1507 It is my mistake. If you look at the list of claimed bounties, you'll see that I bolded Razetime's claim, which is in fact 100. I'll see if I can get a mod to up that. Otherwise I'll bounty something else of yours.
@rak1507 Because bags doesn't use its operand; it looks at its operand, and since ~ is actually a cover function, bags doesn't recognise it. 'abc'`~⌂bags'abd' should work.
<Kiscica> one thing i like about k is that it doesn't try to be everything to everyone. but the choices arthur made were pretty internally consistent and practical
@Wezl imo having easy custom dialects would be very beneficial to APL. Solves the issue of not being able to define custom primitives, and, if done right, means you could have only one common base to optimize to have everyone benefit. (note: a couple hours ago watched Guy Steele's "Growing a Language", am biased)
well, point still stands; there are too many choices with no clear answer in a given APL impl that having them chosen for you is bound to be suboptimal; hence, custom dialects
@Adám you also have Extended & Prime. You don't program anything other than APL, so that's as close as you're gonna get without having a pretty slow end result
<Kiscica> everyone should write a lisp and an apl in their life, regardless od what kind of programmer they are. or even if they aren't a programmer. <-- opinionated statement