The APL Orchard

Fmbalbuena

12:42 AM

@Adám are you Creator of APL?

Adám

What? Nobody owns APL.

Fmbalbuena

@Adám Who created APL?

Fawn Locke

Kenneth E. Iverson

Kenneth Eugene Iverson (17 December 1920 – 19 October 2004) was a Canadian computer scientist noted for the development of the programming language APL. He was honored with the Turing Award in 1979 "for his pioneering effort in programming languages and mathematical notation resulting in what the computing field now knows as APL; for his contributions to the implementation of interactive systems, to educational uses of APL, and to programming language theory and practice". == Life == Ken Iverson was born on 17 December 1920 near Camrose, a town in central Alberta, Canada. His parents were farmers...

As Adám stated, modern APL isn't owned by anyone.

Adám

APL was never owned by anyone.

Fmbalbuena

who owned modern APL?

stopped adding updates?

Adám

12:46 AM

@Fmbalbuena Not sure what you're asking.

Fmbalbuena

or features?

Adám

I still don't get it. Are you asking "who stopped adding features?" or "has the addition of features to APL stopped?"

Fmbalbuena

has the addition of features to APL stopped?

Adám

No, absolutely not.

Fmbalbuena

Whats the Date of last update?

dzaima

12:49 AM

there is no single APL.

Fmbalbuena

pjC█d┐│↑Ψ{⁽○σ#d{§⁰!∆*κ‽Τ!P¤‚↕⁴‘

Adám

@Fmbalbuena You can find that on the page for each implementation on APL Wiki.

@Fmbalbuena Wrong room?

dzaima

no, that's just my "about me"

Adám

Ah, yes, I knew I'd seen it somewhere.

Fmbalbuena

@Adám but what?

dzaima

12:52 AM

@Fmbalbuena There's Dyalog APL. There's GNU APL. There's NARS2000. (then there's dzaima/APL. Then Kap, April, and some more probably)

Fmbalbuena

@dzaima Date?

dzaima

each has it's own features, and thus updates

Adám

@Fmbalbuena Look at the page for each one on APL Wiki, e.g. apl.wiki/Dyalog_APL and apl.wiki/APL%2BWin etc.

Fawn Locke

Kap, April & Dyalog are the most actively updated.

dzaima

here are at least 25 APLs

Fmbalbuena

12:55 AM

@dzaima What about this?

dzaima

@Fmbalbuena Those aren't really APLs, but are just inspired by some features of it. (and APL isn't an esolang)

Fmbalbuena

@Adám Split text by delimiter? in APL?

Fawn Locke

@Fmbalbuena aplcart.info/?q=split%20text%20#

Adám

@Fmbalbuena aplcart.info?q=split text by delimiter

Fmbalbuena

Please show using TryAPL bot

Fawn Locke

12:58 AM

⋄ ','(≠⊆⊢)'comma,seperated,string'

TryAPL

@FawnLocke
┌─────┬─────────┬──────┐
│comma│seperated│string│
└─────┴─────────┴──────┘

dzaima

@dzaima (fwiw, that's SOGL code)

Fmbalbuena

ok so i have to gtg. i'm back tomorrow.

Fawn Locke

Cya :)

Adám

○/

Fawn Locke

1:05 AM

Adám, I know you can define new primitives in Dyalog. dx and dyalog-apl-extended are proof of that. But is it possible to redefine the primitives? I see there's a definition for iota in apl-extended, so most likely. Just curious what file to look at so I can reproduce it

Adám

@FawnLocke "you can define new primitives in Dyalog" is a false premise. No, you cannot.

Fawn Locke

Oh? Huh. How do dyalog-apl-extended/dx work then, is it just fancy eval?

Adám

Text substitution and then plain eval.

Fawn Locke

Gotcha, thanks

elliptic00

8:01 AM

H ← {÷1+∘.+⍨⍳⍵}

2

Dyalog '11: Rational Numbers

►

I just watch this video, I got different matrix from the video in RIDE..

if you forward the video to 6:12, then you can see the function,

The matrix is different from my output in RIDE.. does any one know why?

0.3333333333 0.25         0.2          0.1666666667 0.1428571429
0.25         0.2          0.1666666667 0.1428571429 0.125
0.2          0.1666666667 0.1428571429 0.125        0.1111111111
0.1666666667 0.1428571429 0.125        0.1111111111 0.1
0.1428571429 0.125        0.1111111111 0.1          0.09090909091

here is my output in RIDE..

H 5

from the video, they use window XP... which is pretty old MS window.. but I assume the Daylog APL compile should not be different ?

 H ← {÷1+∘.+⍨⍳⍵}

milia

@elliptic00 this is weird, I'm getting the same results with tryapl.org:

H ← {÷1+∘.+⍨⍳⍵}

H 2
0.3333333333 0.25
0.25 0.2

H 2
0.3333333333 0.25
0.25 0.2

H 2
0.3333333333 0.25
0.25 0.2

ooff, sorry for the bad formatting

Could you please run two three times H 2 and paste the results here?

@elliptic00

elliptic00

I think the speaker is Roger Hui, he just pass away not long ago in Vancouver here,

H 2
0.3333333333 0.25
0.25         0.2
      H 2
0.3333333333 0.25
0.25         0.2

@milia, press CTRL-k , then you will get good format..

after you paste your code, CTRL-k

milia

8:23 AM

@elliptic00 thanks for the formatting tip. Yeah, I've learned about Roger a few days after his passing and I've been following his work closely since I started learning APL/J.

So in order to understand your former question...

You're asking why you're getting different results than those from the video ?

elliptic00

Yep

milia

I guess you're right, maybe something has changed in the interpreter. I wouldn't know for sure. Maybe @Adám knows.

elliptic00

Yep.. We can ask Adam when he is online.. from the screen, I can see the video is pretty old video..

@milia, thx..:)

milia

Things change over the years I suppose :) I'd trust the results of the current interpreter, and then maybe do the calculations by hand on say a 2x2 matrix

or try a python script

to confirm/falsify the current Dyalog interpreter's results

my guess, python's results will match current Dyalog APL's

if you do such a test, lmk, I'm very curious :)

elliptic00

Sure.. I will let you know...

milia

8:31 AM

cool:D

RGS

@elliptic00 and @milia the difference is that Roger Hui always used ⎕IO ← 0 :)

⋄ ⎕IO ← 0 ⋄ H ← {÷1+∘.+⍨⍳⍵} ⋄ H 2

TryAPL

@RGS
1   0.5
0.5 0.3333333333

elliptic00

what is ⎕IO ← 0 ?

RGS

IO stands for Index Origin, it's a system variable that controls, well, the origin of the index :P It determines whether indexing starts at 0 or 1, and it influences primitives like ⍳:

2

⋄ ⎕IO ← 0 ⋄ ⍳5 ⋄ ⎕IO ← 1 ⋄ ⍳5

TryAPL

@RGS
0 1 2 3 4
1 2 3 4 5

elliptic00

8:43 AM

LOL.. i c..

RGS

By default, it's ⎕IO ← 1; however, some people prefer ⎕IO ← 0. To be 100% safe, H could be written as H ← {⎕IO ← 0 ⋄ ÷1+∘.+⍨⍳⍵}.

elliptic00

@RGS, thx for clarifying that for someone like me are new to APL...

RGS

Absolutely! I don't always hang here, but I love helping & teaching

milia

9:21 AM

@elliptic00 yeah, what @RGS said. Pretty much with ` ⎕IO ← 0` you define that your array index will start from 0, as in C, python, Java, etc. If you do it like this ` ⎕IO ← 1`, you're doing it as in Fortran. So yeah, it makes total sense to be getting different results.

elliptic00

Yep, thx for clarifying that.

milia

I think I prefer ` ⎕IO ← 0` too. The Fortran-way seems so unnatural to me nowdays, but even in modern Fortran (eg. F95) I think there are ways to make that index start from 0, as is the norm nowdays.

sure, anytime :)

cool question

xpqz

@RGS looking at your custom.css, I can see no trace of it serving a different font? Does that mean that MDAPL is also problematic in chromium browsers?

Adám

9:50 AM

@milia Is there a reason you put a space at the beginning of your inline code? It doesn't work with SE Chat's markdown: ` code` vs code

milia

@Adám I must have missed that. Most likely I copied a space character just before the APL code as well. Nice catch.

yeah, I guess there isn't some feature like code here ? or is there and I'm not aware of it ?

hm, backticks dissapeared around "code"

RGS

@xpqz Riiiiight, I thought you wanted to see how to specify custom CSS. I have to check those font issues, then :O web dev is so complicated :p

Adám

@milia Not sure what you're asking.

xpqz

@RGS Yeah. Every time I have to do any css I die a little inside.

milia

@Adám in discord if you put backticks around something, it creates a little boxy thing around it.

Adám

9:58 AM

Well, here too does it render without further markdown, and in monospace, but it doesn't change the background colour.

It is much more like an unstyled HTML <code> element.

Have a look at apl.wiki/APL_Orchard#Features

I use a user style sheet (via Stylus) that reformats some things:

milia

@Adám thank you, reading now!

xpqz

10:26 AM

@Adám I've updated the custom css to include the font you suggested. At least, I think I have. Not sure how I can verify it's correct.

Adám

The real way of testing it is to use a device without the font installed.

milia

10:57 AM

@Adám I'm writing a short Google Doc intro to APL, language, inventor, community, try-apl, to share it with my colleagues. I've noticed that some APL symbols such as the tally don't show properly there. Is there a workaround ?

It looks like this in the docs: ≢

Adám

@milia Yes, format the APL symbols using a font that has them: apl.wiki/Fonts

@milia Well, you just copied the symbol, and on my machine, that looks fine, but I know the issue, and it looks like ≡/ by you.

milia

@Adám spot on, on how it looks on my machine. Will check if I can use the Fonts link provided, thanks :D.

so in a normal computer maybe it shows normal, I guess. I'll edit the Doc from my Linux/Windows laptop, because I don't want to mess with the Chrome box.

ty !

Adám

@milia Oh, hm, it is a Google Doc, so you're restricted to the fonts Google supplies…

milia

oh

okay...

then I guess I'll have to take the LaTex route :)

will send it to you as well if you'd like, when done. I'd appreciate any comments or corrections.

Adám

@milia Not necessarily. Google Docs does let you choose font. Choose Roboto Mono

milia

11:05 AM

Oh neat, will try, ty

Adám

milia

hm

not sure what's going on on my doc, but I'm getting even with Roboto Mono the same result as Arial

I'll look into it later, from a different pc too, if it matters

ty

Adám

11:17 AM

@milia Go to "More fonts" and Fira Code.

milia

that did it. Thx !!

Rudi Angela

12:11 PM

Hi all, beginner question here: I have a matrix to which I want to add a row number column. Is there a one-liner for that?

Adám

@RudiAngela (⍳≢m),m – also, welcome!

Kamila Szewczyk

@xpqz you can always follow the ways of Lakatos :)

Adám

@RudiAngela I assumed you want the column on the left. Was that correct?

xpqz

@KamilaSzewczyk Attempt to resolve the perceived conflict between Popper's falsificationism and the revolutionary structure of science described by Kuhn?

Kamila Szewczyk

have you read "Proofs and Refutations"?

Rudi Angela

12:15 PM

@Adám @Adám: Thank you! For now any location is ok for the extra column. Now let me try your suggestion.

xpqz

@KamilaSzewczyk No. Should I?

Kamila Szewczyk

@xpqz "Proofs and Refutations" is a dialogue between students

you can see it for yourself:
https://palaiologos.rocks/library/Imre%20Lakatos%3b%20Proofs%20and%20Refutations.pdf

Adám

@RudiAngela You can try it right here: ⋄ m←4 3⍴⎕A ⋄ (⍳≢m),m

TryAPL

@Adám
1 ABC
2 DEF
3 GHI
4 JKL

Rudi Angela

@Adám Cool!

Kamila Szewczyk

12:17 PM

I think that this book is a compulsory lecture for all teachers

xpqz

@KamilaSzewczyk Nice. It reads almost like a stage play

Kamila Szewczyk

do you dare to write a stage play about APL :)

PyGamer0

CMC: Remove adjacent duplicates in a vector. For example: 1 2 3 4 4 3 2 1 2 2 3 4 gives 1 2 3 3 2 1 3 4

Adám

@PyGamer0 What does 1 2 2 2 3 give?

@PyGamer0 dzaima/APL, 13: ⊢⌿⍨1(,∧,⍨)2≠/ Try it online!

Oh wait, this assumes a simple vector. ≢ instead of ≠ for the general case.

xpqz

12:37 PM

@KamilaSzewczyk was thinking more like a soap opera.

Kamila Szewczyk

aw

ever considered awakening your inner artist instead :)?

milia

12:59 PM

@KamilaSzewczyk not surprised to see one of the most important philosophy of science books on this room.

but rather pleased to see it :)

milia

1:31 PM

this is a pretty nice intro in philosophy of science, for anyone interested to do the deep dive and have a great headache after reading it. Great read, but kinda hard. Surely interesting tho. en.wikipedia.org/wiki/What_Is_This_Thing_Called_Science%3F

Adám

2:07 PM

Announcement: Most recordings from Dyalog '21 have now been published.

2

milia

2:22 PM

@Adám regarding next years' Dyalog user meeting ('22), do you know if there could be an option for someone that'd prefer to stay to some other hotel, and have a fee just for attending the talks/workshops?

Adám

@milia Assuming it'll be IRL, then yes, we do allow you to book attendance without lodging.

milia

cool thx !

it'd be nice if it'd be IRL. I see it's only 3h40m far from my place, and seems to have direct flights too.

Adám

Yeah, it'd be very nice. It is close to RGS too.

milia

:)

xpqz

3:37 PM

@Adám did you get a chance to cast your eyes over the cultivation-as-notebook?

xpqz

5:05 PM

J on hacker news: news.ycombinator.com/item?id=29338799

elliptic00

6:29 PM

∘.+⍨ ⍳ 2
2 3
3 4
      1 + ∘.+⍨ ⍳ 2
3 4
4 5
      ÷ 1 + ∘.+⍨ ⍳ 2
0.3333333333 0.25
0.25         0.2

⍝ How the ÷1 apply to the matrix?

1 + (∘.+⍨ ⍳ 2)

1 is add to the matrix

⍝ I'm wondering where ÷1 come from?

dzaima

÷ there is monadic, i.e. reciprocal. It's ÷ (1 + (∘.+⍨ (⍳2)))

elliptic00

8:03 PM

oh.. it is reciprocal of each element of the matrix.. .@dzaima, thx:)

elliptic00

9:44 PM

Dyalog '21: Scripting in Dyalog v18.2

►

Finally, Dyalog makes the APL running as script

when we can do it and use it?

Rudi Angela

10:01 PM

Question: Given a list of integers I want to produce a list of occurrence count per value in the list. The following works, but I would like to pass the list as a right parameter instead of having a reference to the specific list in the expression. Any suggestions? {⍺,¨{⍵ +.= ∘, mylist}¨ ⍵}⍨ ∪mylist

rak1507

the key operator is exactly what you want here, {⍺,≢⍵}⌸

dzaima

alternatively, a literal translation: {u,¨+.=∘⍵¨ u←∪⍵} mylist (or {(⊢,¨+.=∘⍵¨) ∪⍵} by utilizing a train)

rak1507

but if you want to do it with that method, you can pass it as an operand so you'd do something like (⍵ {⍵ +.=∘, ⍺⍺})¨ or use it as a left argument ⍵∘{⍵+.=∘,⍺}¨

dzaima

note that rak's ⌸ example will give a matrix; {⊂⍺,≢⍵}⌸ gives the exact same result as yours

(also, {⊂⍺(≢⍵)}⌸ would be more correct, if your list isn't a simple list)

rak1507

yeah true

Rudi Angela

10:08 PM

Nice suggestions! I'll go try them. Thx!

Transcript for