@rabbitgrowth I don’t know if it’s the fastest but it probably should have been the most obvious.

How about ⊃∘⌽⍸?

⊃∘⌽⍸ feels much more obvious than ⌈/⍸ to me because normally you wouldn't take the max of a list of numbers when you know for sure that it's increasing

⎕DR triggers a compaction, right?

I vaguely recall there being a way to get the current DR without triggering gc and friends.

@B.Wilson this i-beam?

181⌶ That's it!

I've added some keywords for 181⌶ to APLcart, so you can find it with "dr without gc"

by the way, do ibeam numbers make any sense or are they just picked at random

@RubenVerg Some make sense, but some are just grouped by number, e.g. 517n.

Maybe we should add a section to apl.wiki/Mnemonics for Dyalog I-beams?

the last section on there is really confusing

how is remembering that + has default left -2i*Im(x) helpful

I suppose × with left (abs(x) + [x = 0])^-1 is correct but uh

why

@Adám okay some of these are quite funny

@RubenVerg You're right, it is a bit of a stretch.

what's the history for "nub" being used to mean filtering unique elements?

as in nub sieve

Haskell calls its Unique function nub

though the docs don't say anything else about it, just the meaning of the word

@Adám Rocking it hard as usual. Günaydin.

@RubenVerg Might just have been Ken, from archaic English nub = essence.

I think we need a list of things he named.

too many to remember!

@NeRoboto Hi there, interested in APL?

Does anyone understand why on page 108 of Aaron Hsu's dissertation the "Before" tree has 18 nodes, but the "before" vectors on the next page only have 16 items each?

I tried to map the node IDs (i) onto the "Before" tree based on the parent (p) and type (t) vectors, and this is what I got:

15F                    0F
  ├─────┬───┐           │
 3G    8A 10E          1B
  ├──┐  │   ├───┬───┐   │
 ?A 5B 9N 11A 13P  ?A  2V
  │  │      │       │
 4V 6A    12N     14V
     │
    7N

It looks like the A (which I assume stands for "array") nodes marked ? are omitted from the vectors. Is this intentional?

more mindset than memorization i'd say

lots of learning, but it's not learning by heart mostly

@NeRoboto Not really. There are actually very few language elements, and they mostly abide by a very consistent call syntax, so once you know the syntax of four, you know how they almost all of them work. And as to the symbols, there are strong mnemonics.

@rabbitgrowth I’ve been through it in the past but would have to get back up to speed to comment. I did find some errors in some of the pictures though.

Actually the picture on 108 is exactly the one that was wrong. He said he posted a correction somewhere but I’m not sure where.