I am currently building an Alexa Skill using my own .NET-Backend, which should return an AlexaTextList with dynamic contents from my service. The request by Alexa looks the following:
{
"version": "1.0",
"session": {
"new": false,
"sessionId": "amzn1.echo-api.session.d5f8d9ee-bdd0-4c70-96...
]aplcart -??
───────────────────────────────────────────────────────────────────────────────
]TOOLS.APLcart
Search APLcart for how to do things
]APLcart [<terms>] [[-list[=n]] [-raw[=all]] | [-url]] [-browser] [-window] [-theme=b|w]
<terms> space separated search terms (use /term/ for regex)
-list list all results (not valid with -url; default: 10)
-list=<n> list no more than <n> results
-raw return raw two-column matrix (not valid with -url)
-raw=all return raw eight-column matrix
C"M"C: golf a chess engine in APL; use d2d4 algebraic style for movements, add some sort of board display, promotion to queen, no en passant, no castling, the engine is valid if it can beat a minmax depth=2 engine most of the time.
bonus points if the resulting program is smaller than a kilobyte :p
@rak1507 Same problem as with everything else. It'll be in the release announcements, and we're adding a "Getting Started" entry to the Help menu. This takes you to a webpage that mentions APLcart.
]APLcart remove blanks -l=5 ⍝ list first five results for "remove blanks"
]APLcart /highest|lowest/ ⍝ list up to 10 results for "highest" or "lowest"
]APLcart -b ⍝ open APLcart in browser
]APLcart -w -t=b ⍝ open black APLcart GUI
]APLcart -u ⍝ internet address for APLcart
]table←APLcart -r=all ⍝ database dump
Sure, I wasn't counting that. These are APLcart specific needs: Check modifier conflicts: 5 Construct URL: 5 Open browser/window: 12 Limit and messages: 8 Header and footer: 10 Format output: 10
If you were testing all window sizes idk what the time complexity is but I feel like it'd be pretty bad because it'll have to 'loop' through the vector several times
so you take primes from 0 to 1000000, store them in p, then scan it all with + and store it in a
and remove everything greater than 1 million, then take outer product with subtraction, ravel it, take the absolute value and take unique values and store it again in a for some reason
I'd say leading axis functions use their n dimensional argument as a vector of (n-1) dimensional arrays, while trailing axis functions use their n dimensional argument as an (n-1) dimensional array of vectors.
'not a single person who I consider to be a real prospect for writing substantial new application code in APL has brought up a single one of these issues as a reason why they might not want to adopt APL' survivors bias 'no one who plans to use APL to write an application has these complaints!'
@Razetime 1) it doesn't have bridges to SE and TA (and even if it did, chatting from it would look ugly, like the current IRC bridge does; also on my todo lists is to finish my SE↔Matrix bridge, but that still won't help with TA); 2) its GUI options suck
@dzaima I've been working on it some, though. I have half of a post walking through how the compiler would interpret (?) a short program, so maybe I'll see if I can finish that and post it.
Progress on the Idris draft is kind of slow because parsing in a pure functional language is annoying, and because I think anything able to demonstrate stuff at all will require closures, which are difficult without mutable data.
@dzaima I find I pretty much need ADTs to reason through all the data structures, so I wanted a language that supports and checks them well.
@rak1507 There are parser combinators, but then I have to learn how parser combinators work, which comes too close to yak-shaving territory for comfort.
Although if you know them and are willing to help then it probably wouldn't be too hard.
Yes, I find that a bit harder but it also feels a bit less important right now -- I want to be able to read tacit that others (incl APLCart) write, but for code I write myself I'm usually ok with dfns.
@user5314949 If you want to participate here, email [email protected]
@xpqz Fair enough, but I'm telling you, once trains become natural to you, they'll become a tool of thought. You'll start "seeing" forks and overs etc.
@KamilaSzewczyk yes I used a matrix maths method with bignums but I'm going to redo it at some point
@xpqz I agree with adam, once you get used to it, certainly short tacit functions become a lot more 'simple' than the equivalent dfn (⊣÷⍥!-) vs {(!⍺)÷(!⍺-⍵)}
I am currently building an Alexa Skill using my own .NET-Backend, which should return an AlexaTextList with dynamic contents from my service. The request by Alexa looks the following:
{
"version": "1.0",
"session": {
"new": false,
"sessionId": "amzn1.echo-api.session.d5f8d9ee-bdd0-4c70-96...
