@Adám managed to solve that question from yesterday, I'm currently at '.!'[?6⍴2]∘{⎕←'Sending 5, 100-byte ICMP Echos, timeout is 2 seconds:',⎕TC[2],⍺[⍳⍵]⊣⎕DL 2}¨⍳6. Any clear golfs anyone can see?

@J.Sallé Run ⎕ML←3 and change ⎕TC[2], to 2↓⎕TC,

Will do

@J.Sallé ⍺[⍳⍵] → ⍵↑⍺

Oh yeah I missed that

@J.Sallé If you're allowed trailing spaces, change ⎕ML back to 1, and just '.!'[?6⍴2]∘{⎕←↑'Sending 5, 100-byte ICMP Echos, timeout is 2 seconds:'(⍵↑⍺)⊣⎕DL 2}¨⍳6

Goddamnit I just noticed I need to print a different string after all of those D:

I'll see if I can have trailing spaces

OP doesn't specify so I assume it's allowed?

@J.Sallé If you're not allowed trailing spaces: '.!'[?6⍴2]∘{⎕←⍵↑⍺⊣⎕←'Sending 5, 100-byte ICMP Echos, timeout is 2 seconds:'⊣⎕DL 2}¨⍳6

@Adám I put a comment about the trailing spaces, but I'll assume for now that it's allowed

aight.

@Adám notably a few of my solutions to challenges on anagol were too slow in Burlesque because the shortest way usually calculates "too much".

(e.g. rosettacode.org/wiki/Find_the_missing_permutation#Burlesque)

Okay, final code is '.!'[?5⍴2]∘{⍵≤5:⎕←↑'Sending 5, 100-byte ICMP Echos, timeout is 2 seconds:'(⍵↑⍺)⊣⎕DL 2⋄⎕←'Success rate is ',⍕(100×d÷5),'percent (',(⍕d←+/'!'⍷⍺),'/5)'}¨⍳7

@mroman Your question was: why use APL? Using APL dosn't mean "golf your code". True, sometimes the shortest approach is terribly inefficient, but that doesn't detract from short code being a plus as long as it stays readable, of course. Usually, APL doesn't do constraint-based computations (there's only one exception to that, and you have to be explicit about it if you want it).

@mroman Rather a team of geniuses can optimise low-level algorithms in the interpreter, and then everybody who uses the interpreter gains a speed-up at no further cost to them.

but you can offer a lot of those things using libraries in languages.

For example extension functions in C# (LINQ).

I mean.. the benefit of using a language would have to outweight the cost of learning the language.

either that or it has a huge DSL character.

(such as some languages are mainly used because they ship with a lot of utility in some area.... like statistics)

I do agree that short code vs long code may have some readability advantage in the sense that you can pack more information into a single line of code.

Reading nested loops across a hundred lines of code with dozens of function call is pretty much less readable than when you have a high-level concept for this stuff in your language.

on the other hand working on such a language and getting paid for it sounds like fun :D.

Is WC a complete win32 API interface?

if not, how would I bootstrap external dependencies like that into APL?

nevermind

⎕nq can help

nq or na

Probably ⎕na and I misremembered

nq looks like system event queuing, so that may be right

na is the FFI

I was looking for that when you said nq

well, thanks for both. There goes all of the rest of my free time :/