The APL Orchard

ngn

17:04

@Marshall yay (belated). and tbh i don't understand why you keep comparing the apl wiki with wikipedia

@RomillyCocking what do you think about "projection"?

dzaima

@ngn because it is the same idea. It has the same goal of describing things (more specifically about APL, but the idea is the same)

@ngn definitely way harder to "guess" what it means

rak1507

I've never heard that term before

ngn

@dzaima three important differences: [0] the apl wiki is (not suprisingly) about apl. wikipedia is about all walks of life. [1] the apl wiki is controlled by dyalog, a commercial company with commercial interests. wikipedia is controlled by a US non-profit. [2] wikipedia is incomparably larger

rak1507

is the APL wiki controlled by dyalog?

Adám

No. It is a normal wiki like Wikipedia.

Elias Mårtenson

17:17

@rak1507 Would Dyalog take down content that is critical of the Dyalog APL product?

Adám

No. There is no censorship or even review from Dyalog's side.

rak1507

aplwiki.com/wiki/Defined_function_(traditional) can any page get more critical of Dyalog APL than this one? (joking ofc)

Adám

I'd happily contribute to an article titled "Syntactic issues in Dyalog APL".

ngn

@rak1507 whois says "redacted for privacy" but the contact emails are @ .it and @ .co.uk

rak1507

Owning the domain ≠ controlling the content to me

Adám

17:22

I think Kai Jäger controls the domain.

ngn

@rak1507 well true, but i still wouldn't even try writing anything important there

dzaima

@ngn 0) yeah, obviously. Doesn't change much other than what's on-topic. 1) That should obviously not matter, especially so when discussing changes publicly. 2) again, shouldn't change much.

rak1507

@ngn Why not?

Adám

Marshall has unlimited rights on the wiki, and he's not associated with Dyalog any more.

ngn

@rak1507 waste of time, they'll edit it

Adám

17:23

@ngn You could give it a try at least.

rak1507

Don't know that until you try it

dzaima

@ngn you don't write "important" things in a wiki

ngn

@rak1507 how hard do you think it would be for me to start my own wiki :)

rak1507

@ngn It would probably take about 3 lines of K knowing you

ngn

@rak1507 you don't know me well enough - 0 lines, as there's no point doing it :)

rak1507

17:26

lol

dzaima

@ngn if they do, there's SE (and Matrix if SE fails) for you to publicly note any and all wrongdoings

ngn

but we're getting distracted.. if the apl wiki follows wikipedia's rules, why isn't it a part of wikipedia?

dzaima

@ngn because of the APL scope. That's, like, the only difference

rak1507

Huh?

Adám

@ngn That axiom is a mistake. So you make mistaken conclusions.

ngn

17:28

@Adám marshall seems to follow that axiom. is he mistaken?

Razetime

@Adám What's the best way to mention the userscript?

dzaima

@ngn there's like one rule that's different between APL wiki and wikipedia.org - scope.

Adám

@Razetime The SE chat one? I'd say in the APL Orchard article and under Running APL.

And scope is super important. That's what causes there to be more than one wiki in the world.

Razetime

Ok sure

dzaima

@Adám I think SE userscripts are best left to the single SE page

ngn

17:32

@dzaima surely "apl" is a subset of "everything notable" :)

Adám

@dzaima Possibly.

@ngn True, but aspects and details of APL are not notable for Wikipedia.

dzaima

@ngn APL-notable and world-notable are very different scales. dzaima/APL definitely doesn't deserve a wikipedia.org page

ngn

@Adám @dzaima qed. they don't follow the same rules.

dzaima

@ngn they follow most (i.e. approximately all except scope) of the same rules.

Adám

6 mins ago, by Adám

@ngn That axiom is a mistake. So you make mistaken conclusions.

@dzaima That's not true either.

ngn

17:35

@Adám i replied to that. you don't have to repeat it.

Adám

@ngn Oh, so you meant to reverse position and agree with me?

dzaima

@Adám it was a purposefully written argument by contradiction

ngn

wat

Adám

Got me confused, here.

dzaima

@ngn I assumed you wrote the message as a "it isn't part of wikipedia, therefore it does not follow wikipedia's rules"

Adám

17:37

@dzaima APL Wiki also has different rules for verifiability too.

dzaima

@Adám ah right. Still, the changes didn't matter as far as the conversation went

Adám

Nope, you're right.

ngn

i think it should have different rules. i was asking marshall why he assumes it has the sames rules as wikipedia.

dzaima

@ngn what rules should be different?

Adám

@ngn Which was a question based on the mistaken assumption that he assumes so, no?

ngn

17:39

@dzaima this for instance

dzaima

@ngn well, then it wouldn't be a wikipedia anymore

Adám

"A wikipedia"‽

ngn

@dzaima wiki to wikipedia is like car to carpet

Adám

OK, that's just nonsense.

rak1507

Wikipedia is an example of a wiki in my mind

dzaima

17:41

@Adám i kind of have a mental equivalence between encyclopedia and wiki/wikipedia

Adám

Wiki, then. Wikipedia is a wiki (or a collection of wikis, if you look at each language as a separate entity) with a specific scope.

dzaima

@rak1507 agreed

@dzaima s/wikipedia/wiki

rak1507

What is this discussion about anyway I've kinda lost track because of all the small irrelevant details being discussed

ngn

@dzaima wikipedia is a huge community with its own rules. mediawiki is just the php application wikipedia runs on. mediawiki is an impl of a wiki (a quickly-editable web page). there are many other wikis.

dzaima

@rak1507 tradfn naming → this → discussion about wiki rules

rak1507

17:45

What's the difference between describing use and specifying use?

Oh right, I think I see what he means

dzaima

@rak1507 former is saying that defined functions are often tradfns. Latter is saying that we should use tradfn or del-function for them everywhere.

@ngn I define "wiki" to be much more than quickly editable web page (including being descriptivist).

Adám

Wiki

A wiki ( (listen) WIK-ee) is a hypertext publication collaboratively edited and managed by its own audience directly using a web browser. A typical wiki contains multiple pages for the subjects or scope of the project and could be either open to the public or limited to use within an organization for maintaining its internal knowledge base. Wikis are enabled by wiki software, otherwise known as wiki engines. A wiki engine, being a form of a content management system, differs from other web-based systems such as blog software, in that the content is created without any defined owner or leader, and...

Romilly Cocking

@ngn Projection is technically correct, but I think some people will not be familiar with the term or guess its meaning.

@dzaima I agree with @Adam - I think it's Ward who gets to define what a wiki is.

ngn

18:08

@RomillyCocking but ward agrees with me (wiki≠wikipedia), so.. :) i'm old enough to remember the c2

Romilly Cocking

@ngn I agree with Adam about who gets to define wiki. I agree with you about wiki ≠ wikipedia. And fwiw I knew Ward before he invented the wiki.

ngn

wow, cool!

Romilly Cocking

My sandwhich of being an APLer had a thick slice of being a Smalltalker in the middle.

So I first met SUnit (precursor to JUNit and friends) in or around 1989. (Kent Beck, whom I also know, is not sure what year he sent me a copy.)

ngn

re "projections" - first i didn't like it for the reasons everybody mentioned, but when i think about it, it does fit nicely with k's unification of syntax for func apllication / indexing / dict lookup. it's probably a less fitting term for the original apl.

dzaima

18:24

@ngn moreover, afaict it's making up another term when there are good alternatives

Marshall

@ngn APL Wiki's content guidelines are here and explicitly say that content policy is based on Wikipedia's.

It includes all the changes to Wikipedia's rules I've seen fit to introduce: the only real changes are the additional paths to article notability and verifiability. Everything else is more like the subject-matter-specific clarifications that Wikipedia itself uses.

Given that I probably wrote more than half of the content on APL Wiki, I think I'm something of an authority on what it is.

ngn

@Marshall because you wrote them :)

ok, fair enough.. the apl wiki follows wikipedia's guidelines, so be it

Adám

\○/

ngn

@Adám it doesn't depend on me :)

it probably doesn't matter much anyway

Marshall

I'm not just blindly following Wikipedia because it's the biggest, by the way. I think Wikipedia currently has some pretty serious problems. But the rules about content and style and editing were formed when Wikipedia was much smaller (only a few orders of magnitude larger than APL Wiki, not a lot of them). At that time it was clearly outcompeting other wikis both in editor participation and readership.

Reading APL Wiki's content policy, I almost always find it to be well-reasoned and applicable to APL Wiki.

ngn

18:32

@Marshall did you write that too? :P

Marshall

@ngn No: as you can see I've made extremely few contributions to Wikipedia.

ngn

@Marshall i mean "apl wiki's content policy"

Marshall

@ngn Oh, that first "APL Wiki" should be "Wikipedia". Wikipedia's content policy. APL Wiki's content policy, while applicable to APL Wiki, is not really that well-reasoned.

Other than the parts about deferring to Wikipedia.

There's also another benefit in staying close to Wikipedia in that readers and editors who are familiar with WP will know what to expect. If we used, say, infoboxes in a completely different way from WP, it would be harder to navigate. I think changing the way the text works would have a similar but harder to describe impact.

ngn

makes sense

rak1507

23:05

cmc: given a number n return the number of permutations of 1 through n (inclusive) where for all i 1 ≤ i ≤ n perm[i] is divisible by i or i is divisible by perm[i]

(leetcode.com/explore/challenge/card/…)

Adám

Maybe I'm just ignorant, but what does perm[i] mean?

rak1507

the ith element of the permutation

Adám

Oh.

rak1507

interested if there's a nicer way of checking if a divides b or b divides a than what I have

Adám

0∊|,|⍨?

Bubbler

23:11

I was thinking of 0=|×|⍨

Adám

Or just 0=|⌊|⍨

rak1507

I have a 3 char solution but it takes 5 bc I have it in a train in parens

Adám

∨∊,? I'm not sure, though.

rak1507

∨≡⌊

Adám

∨=⌊ is more array-oriented.

rak1507

23:13

∨≡⌊ checks if the entire permutation satisfies that

Adám

Ah.

So +/dfns.pmat(∨≡⌊)⍤1⍳?

rak1507

yep

Adám

Makes me wonder what's usually needed most, leading or trailing axis agreement.

So I have this idea of a RISC-APL where all functions are defined with he maximum possible rank (and depth), and then you use the rank and depth operators to apply them.

But maybe, if it could be constructed, a language where the functions apply on atoms by default, and then you could raise the "level" of application, would be better.

It'd require "scalar" functions to not penetrate boxes, just like J does it.

There'd be no ≡, only =⍥⊂ but instead of ⊂, it would be a count of levels to enclose.

Transcript for