The Nineteenth Byte

4:07 AM

Anyone willing to help a poor me:

APL=code=>{
  r=(
		_=>{
      return new Promise(
        (resolve, reject)=>{
          x=new XMLHttpRequest
          x.onreadystatechange=_=>{
            if(this.readyState==4&&status==200){
              resolve(this.responseText);
					  }
          }
          x.onerror=reject
          x.open("POST","https://tryapl.org/Exec")
          x.setRequestHeader("Content-Type","application/json;charset=utf-8")
          x.send(JSON.stringify([,,,"⎕JSON⍠'HighRank' 'Split'⊢"+code]))
        }
      )

You need to return that last line in the arrow function

return r.then(...)

I just get the unresolved promise as result. I console reads Promise { <state>: "pending" }

You're going to need to use async functions or a callback

What do you do with the output of the APL function?

I want to return the response.

the r.text thing is just nonsense in my attempts at getting anything back.

You can just await APL(...) if it's in the console or an async function

4:11 AM

Yeah, I tried. It just goes silent then.

In reality, I'd run JSON.parse(JSON.parse(resonse)[3])

@RedwolfPrograms Can I not put that await inside the APL function?

2

It would need to be an async function, but then it would return a promise too

It needs to be this.status not status

Actually since you're using an arrow function it needs to be x.status and x.readyState

And x.responseText

Here, this works:

APL=code=>{
  r=(
		_=>{
      return new Promise(
        (resolve, reject)=>{
          x=new XMLHttpRequest
          x.onreadystatechange=_=>{
            if(x.readyState==4&&x.status==200){
              resolve(x.responseText);
					  }
          }
          x.onerror=reject
          x.open("POST","https://tryapl.org/Exec")
          x.setRequestHeader("Content-Type","application/json;charset=utf-8")
          x.send(JSON.stringify([,,,"⎕JSON⍠'HighRank' 'Split'⊢"+code]))
        }
      )

The .text was unnecessary, it just caused the promise to return undefined. The then wasn't even necessary at that point.

Wow. Thank you.

No problem, JS is weird :p

@RedwolfPrograms Should I not be able to wrap this all up then?

APL=async(code)=>{
TryAPL=code=>{
  r=(
		_=>{
      return new Promise(
        (resolve, reject)=>{
          x=new XMLHttpRequest
          x.onreadystatechange=_=>{
            if(x.readyState==4&&x.status==200){
              resolve(x.responseText);
					  }
          }
          x.onerror=reject
          x.open("POST","https://tryapl.org/Exec")
          x.setRequestHeader("Content-Type","application/json;charset=utf-8")
          x.send(JSON.stringify([,,,"⎕JSON⍠'HighRank' 'Split'⊢"+code]))

(I'll fix the indentation later)

4:28 AM

Is that missing some code at the end?

No, but I think I got it. This works (but maybe can be simplified):

APL=await (async(code)=>{
TryAPL=code=>{
  r=(
		_=>{
      return new Promise(
        (resolve, reject)=>{
          x=new XMLHttpRequest
          x.onreadystatechange=_=>{
            if(x.readyState==4&&x.status==200){
              resolve(x.responseText);
					  }
          }
          x.onerror=reject
          x.open("POST","https://tryapl.org/Exec")
          x.setRequestHeader("Content-Type","application/json;charset=utf-8")
          x.send(JSON.stringify([,,,"⎕JSON⍠'HighRank' 'Split'⊢"+code]))

I think you could remove the whole r thing and just return the promise from TryAPL

@RedwolfPrograms Thank you again. Lots of clean-up later:

APL=await (async(code)=>{
  return JSON.parse(JSON.parse(await(code=>{
    return new Promise(
      (k,x)=>{
        with(new XMLHttpRequest){
          onreadystatechange=_=>{if(readyState==4&&status==200)k(responseText)}
          open("POST","https://tryapl.org/Exec")
          setRequestHeader("Content-Type","application/json;charset=utf-8")
          send(JSON.stringify([,,,"⎕JSON⍠'HighRank' 'Split'⊢"+code]))
        }
      }
    )
  })(code))[3])
})
APL("⍳3")

Oh well, the mixture of tabs and spaces is messed up, but other than that…

4:45 AM

Looks good

Hm, that outer await doesn't seem to do what I want.

Oh, I just noticed that was there.

You're going to have to have APL return a promise

Yeah. How do I wrap this so I await the result rather than returning the promise?

You can't just return the result of an async function from a normal one, it just sort of "trickles down" as promises sort of

Not really sure how to explain it, but you'll either need a callback or an await outside the function

Sure, so lets say I go with an await outside the function… Can I then not write another function that has the await?

4:57 AM

Do you mean something like (promise) => await promise?

Yeah.

No, because JS requires that await only be used in async functions, which return promises that would need to be awaited, and so on.

So I actually must use await whenever I call my function? How silly.

In other words, if I'm given a blackbox JS function, I can't know if I need to call it with await.

You could have a function that takes a blackbox function and always needs its result to be awaited, but not one that never does

Razetime

hello

5:03 AM

My idea was to allow inline APL code in JS, as in timesTable=size=>APL(`∘.×⍨⍳${size}`)

Razetime

woah

Unfortunately I don't think there's any way of doing that, as long as there's something like an XMLHttpRequest that requires async :/

Oh well, there went that night ¯\_(⍨)_/¯

Requiring an await isn't that much of an annoyance, unless browser compatibility was a concern

No but that means all calls to such functions would need awaits too.

Can I not just put await this at the top of my script‽ /s

5:10 AM

A different, still kind of annoying, option would be callback functions

I don't understand how that would allow inline usage.

It wouldn't, but I guess it's slightly better than async spam

Either way defeats my goal of seamless integration of APL into JS.

Actually I'm kind of halfway wondering if you could use a while loop to wait for the promise to finish

Let me try something

Never mind, the while loop would block any other code from running. I would try using timeouts but that would also require async :/

!

Neil

12:08 PM

on a side note you need to declare x as local otherwise if you start two requests then they'll reuse the global x and it won't work properly

oh wait I see you do in a later version

but yeah, sync functions expect to run to completion, so there's no wait to make them wait for something to happen

I guess some command-line scripting languages allow you to wait inline for an async call, but that doesn't play nicely with a browser language

I think web workers might be able to do that, but then of course invoking the web worker is still async from the point of view of the page itself

Anush

2:29 PM

I have a list of short strings in python which can be compressed. How can I compress them and store the result in the code so it can then be decompressed by more code?

Maybe based64 encoding would come into it?

It's basically a python kolmogorov complexity question

the default.

Python also has base85. (but I don't understand what are you trying to use base64 for)

3:23 PM

journal.stuffwithstuff.com/2015/02/01/… relevant article

Anush

4:13 PM

@thedefault. I just want to be able to compress the strings and encode them in my code. Base85 looks better.

How would i go about this?

The problem is base85 expands everything...hmm..

Does anyone remember a question where there were multiple R answers trying to outgolf each other? I believe the comments contained "Ping-pong" somewhere.

why compress the strings anyway, unless you're code golfing

@user Browse through either Dominic van Essen's or Giuseppe's recent answers, that's probably a good start

@cairdcoinheringaahing Good idea

4:54 PM

Man, it's tough to write a self-nomination for "Underappreciated answer" without sounding like you're complaining about not getting enough votes :/

4

does your username come from this answer?? codegolf.stackexchange.com/a/118068/95516 @cairdcoinheringaahing

2

@rak1507 Yep :)

ah cool I always wondered what it meant

TBH I've been meaning to fully break down the decompression and see which bits of it are letters and which are dictionary words

5:26 PM

8420 nice :p

I'm hoping I'll get above 8500 before the nominations end

all the questions have already been answered! it's very inconvenient

@rak1507 Which questions? We have 70 unanswered questions

all the questions that aren't impossible :P

They might not be impossible, I managed to knock one out that had been unanswered for four years

Turns out my username consists of three words: caird, coinhering and aahing

5:31 PM

how on earth is 'coinhering' a word

> What does coinhering mean? Present participle of coinhere. (verb)

I swear Google is so helpful

seems slightly weird to have in the dictionary

> Coinhere meaning
(intransitive) To inhere or exist together, as in one substance.

Oh, so it's co-inhere not coin-here

lol

5:32 PM

> Caird definition, a traveling tinker, especially a Romani

So my username means "a traveling tinker, existing together and exclaiming in pleasure"

And it should be pronounced "cared co-in-hear-ing-aah-ing"

4

I always pronounced the second part "coin-her-ing-aah-ing" in my head :p

same

I've gone "coin-herring-aah-ing" (not sure if "herring" = "her-ing")

Do you think it'd be fine to self nominate my answer to hexasweep (the one that was unanswered)? Asking because it already got a 200 rep bounty for being first.

@RedwolfPrograms Depends on the category in my mind. I wouldn't consider a bountied answer to be underappreciated for example

Wild Card could be a good one

Huh, we don't actually have just flat "Best Challenge" and "Best Answer" categories

5:40 PM

I guess that's what upvotes exist for, otherwise they'd be in the unappreciated categories :p

Guess so :P

19 hours ago, by caird coinheringaahing

@mods Could we move the featured from the Call for Categories to the Nominations post for the Best of 2020?

5:51 PM

I was going to self nominate Bot Factory KoTH for underappreciated but I think instead I'm going to sandbox it and do a "part two"

@cairdcoinheringaahing Which answer is this?

@user This one - the Enigma machine question

Oh yeah, that was pretty cool (and it didn't get a ton of upvotes because it's such an old question). I'd say it's underappreciated.

Scrolled back to the top of the answer to upvote and it was already green lol

~~Yeah I'd downvoted it already~~

6:06 PM

@RedwolfPrograms Damn bro, SE must have a weird bug where downvoting an answer gives +10 instead of -2 :P

Oh hey, I just got my second nice answer badge

You're welcome :D

Thanks! :p

CMQ: Do you agree with question votes counting for +10 instead of +5?

Yes

Would this count as a wildcard? It's really nice

6:13 PM

@RedwolfPrograms SO doesn't

I think they should make downvotes count for more, one upvote countering five downvotes doesn't really seem like a good balance

Or maybe multiple downvotes in a short period of time would have an extra penalty?

It works for answers, why not questions?

I mean I guess I can't really disagree with it given that it doubled my rep on five sites :p

@user Personally, I think it would fall under Best Explanation more

I should change my profile picture to one of my cats for the irony

6:22 PM

@RedwolfPrograms Of course, we'd have to be sworn enemies :P

@Razetime I appreciate the nomination for the commenter category, but I would've preferred if it were more than just a single sentence. Even just a quick explanation about why you think I deserve it would be good

ಠ_ಠ And I just realised I can't upvote any of the nomination answers because I technically posted them

Aww look at my cat

New Main Posts

7:43 PM

0

Q: Encoding Binary Tree

Suppose we want to develop a tool for building the smaller decision tree. Given a set of samples, we want to find the decision tree with the minimum number of nodes that correctly classify the samples. All features have only 0 or 1. Further, samples always fall into one of the classes, positive (...

8:41 PM

What was the consensus regarding cleaning up the sandbox? I'm considering making a userscript that could help keep track of our decisions regarding each sandboxed post.

@RedwolfPrograms We want to do it, but every attempt to do so doesn't really do anything

In what way?

@RedwolfPrograms We've had attempts to go through the Sandbox and delete inactive posts, or to "pause" it for a bit to edit and delete old posts, all that. But it's >2000 posts, so it's never had much impact

Here's how I think it should be handled:
1. Keep any posts active in the last month
2. Comment on any old ones to see if the owner still cares or is around
3. Put any that aren't claimed on some page or chatroom (maybe "The Garage Sale" or something)
4. Delete the rest after some amount of time

That would probably remove at least half of them

@RedwolfPrograms Deja vu

8:50 PM

Ew, it's written in <h1>

@RedwolfPrograms You can't see deleted posts right?

No, not yet

There are 7 deleted answers to that with different suggestions for fixing the Sandbox :P

@RedwolfPrograms If you want to open up a meta discussion about that, I'd be more than happy to support it, but I doubt much'll come from it

I think the biggest problem is that it requires a lot of work and a lot of moving parts. Remind the users, delete bad abandoned posts, potentially move good ones somewhere else, etc. In addition to the comment necro'ing. I think a userscript could fix most of that.

Seems interesting that the top posts in the sandbox are all KOTH, I suppose they take the most work to do

8:55 PM

A userscript could hide a lot of that, it couldn't do much to fix it

I disagree. It's pretty easy to click a button and vote on whether or not some sandbox posts are worth keeping if you have five minutes of time. It takes a lot more investment to keep track of the current status, what posts are being looked at, what other's opinions are, which ones needed to be deleted, etc.

I wouldn't mind putting a moratorium on the Sandbox for a week or so, have a massive, organised cleanup event then "unfreeze" it and bump the answers that were on the first couple of pages back to be active

But I think that'll be difficult to do well

I think it'll either have to be like that or a slow ten-questions-a-day-until-it's-done thing, but trying to go at a medium pace won't work

@RedwolfPrograms 10/day = 200 days of dedicated cleanup. I love the moderation side of the site and even I'd be burned out by that

If there's two thousand questions and each one takes two minutes of work to classify and comment/remove, that's 70 man hours of work, tops. Twenty minutes a day spread out over ten people is three weeks.

It doesn't need to be spending hours sitting there classifying things, just classify or comment on a few whenever you have five minutes.

9:00 PM

@RedwolfPrograms It'd never take 2 minutes to remove, you need 3 DV from 20k+ users to get rid of an answer without a mod

I'm not trying to shoot down your ideas, just giving some perspective from past attempts to do a similar thing

But clicking the button only takes seconds, esp. with them already classified

It probably would need a mod to do most of the deletion though

In my experience, I think in order to do anything, we'd need to freeze the Sandbox to prevent new answers/updates from screwing up the cleanup

We only get a few answers a day, right? And those would be ignored either way.

@rak1507 non-code-golfs are generally the most difficult to do

yeah

9:06 PM

@RedwolfPrograms Yeah, but you don't want those buried below edited/classified/updated proposals

True

The Sandbox is 6 years old, it might be a good idea to start a new one

We used to have fairly regular Sandboxes, but that used to be like every 6/9 months, not 6 years

Probably bad idea: What if we had different sandboxes for different types of questions?

@RedwolfPrograms Problems with that: the front page of meta will get very clustered. Which "types" get a sandbox and which don't? Most challenges are code-golf, so that's the main one that'll be used and the other's will (unfortunately) die off eventually

Yeah, plus we'd have to be like "try the sandbox! or the sandbox! or the sandbox! or the sandbox!"

10:13 PM

@RedwolfPrograms that's better than how I mentally pronounce it

How's that?

I always read "carrid coinengineering"

...

@Lyxal shudders

@RedwolfPrograms don't you even think of ...ing me like that buckeroo

10:15 PM

But do you read it as "coin-engineering" or "co-inengineering"?

@cairdcoinheringaahing I've literally told you this twice before

@RedwolfPrograms coin engineering

Q: I need to try to find an algorithm to decode a set of Excel dates

@Lyxal And I shudder every time :P

New Main Posts

0

A former employee once told me he could encode a date in Excel and no one would ever figure it out, "like a needle in a haystack". My response was "Why would someone want to do that?" After his departure with the theft of a large amount of data, I find a set of charts labeled anywhere from from T...

date algorithm encryption vba

The timing on that was perfect lol

and of course carrid coinengineering has already commented :p

We definitely need those ask question warnings added, I'll write the meta post asking for ideas tonight so we can decide on what we'll add/change

Neil

of course, what I need to test my KQ vs K endgame winner is a counter-program whose job it is to stymie my program as far as possible (in that that it should eventually fail because the endgame is a clear win)

Bubbler

10:21 PM

@Lyxal I just read "caird" and don't care the rest

@RedwolfPrograms Just because I nominated you for this doesn't mean I'm not going to give you a run for your money :P

@NewMainPosts One more downvote to DVs, one more close vote till closed :P

> This question has been closed since you loaded the page - please refresh and try your action again.

:(

I tried to vtc and got this

Arnauld beat you to it :P

Smh

Cast a downvote tho so I can VTD :P

10:27 PM

Your welcome

(sic)

Got to love removing VLQ posts :)

How does this sound for a pop-con: create a mathematical expression/equation that evaluates to 69 where the only numeric literal usable is 420

Did you hear that? It's the sound of me hitting the VTC button at lightning speed :P

@Lyxal brilliant

o+o+o+o+o++⍨+⍨×⍨×⍨+⍨o←÷⍨420 APL :P

@cairdcoinheringaahing so you don't think it's a good idea

?

10:40 PM

@Lyxal Bluntly honest: no, not at all

:(

Is that because it uses the funny numbers and you're tired of me slipping them in like vegetables in a pizza?

Personally, the "funny numbers haha 69 420" meme is awful IMO, and a challenge based around it isn't ever going to be fun

@Lyxal Not just you, I've never really found the "funny" numbers that funny. It's a very 14yo kinda humour ("haha sex haha weed"), which I despise, looking back on how I was when I was 14

I think it's more about the irony of the meme itself rather than what the numbers refer to, sort of

Of course, if you can write a Sandbox proposal which isn't actually close-worthy, I wouldn't VTC it just because it uses 2 specific numbers, but I would leave a downvote

10:44 PM

39913 people think it's still funny

Good for them. I don't, so I'd downvote a challenge based around them

I'd only VTC it if it actually should be VTCed tho

Well lucky for you I was only seeing how it would be recieved

I don't actually plan on sandboxing/drafting it

Memes don't really seem to translate well into CGCC challenges from what I've seen

@Lyxal Turn it into a CMC and it'll be perfect :P

@RedwolfPrograms We hate fun

@cairdcoinheringaahing says the one with a sandbox post based on a meme subreddit

10:48 PM

@Lyxal shifty glances

codegolf.stackexchange.com/questions/6043/… I mean...

Most memes don't :p

In all seriousness tho, most memes lack the objectivity we require. Some work, quite a few can inspire challenge ideas, but most need to be adapted too much to still be the original meme

^ My feelings exactly

And of all the SE sites, we probably hate fun the least

10:49 PM

What about board and card games :p

We've had a few doge related challenges in the past IIRC

@RedwolfPrograms No-one likes the rules lawyer in a game, that site's full of them :P

this is kinda sad

Is that new questions?

Yep

2 completely off-topic, one that ~~could work, but needs to be Sandboxed. Of course, the 1-rep user can't Sandbox it because yay for bugs~~ is a duplicate

Custom sidebar warnings will probably decrease blatantly off topic questions by at least 50%, I'd bet

10:53 PM

I just watched a 1 bit rickroll and I'm very happy right now

I mean 'Code Golf Stack Exchange is a site for recreational programming competitions, not general programming questions. Challenges must have an objective scoring criterion, and it is highly recommended to first post proposed challenges in the Sandbox. ' seems pretty obvious to me

@rak1507 But it's kinda easy to ignore

true

It'd be a lot more attention-getting with a big yellow triangle.

Maybe we could put it in all caps with unicode bold :p

make it bold, up the font size by 20, add a loud siren sound effect whenever someone visits the site for the first time, add 27 javascript alerts whenever someone starts to submit a challenge...

10:55 PM

Scrap all that and have a pop-up, 2 window tutorial on how to ask

That's basically what the initial modal is for

Anyway gtg be back in 4 hours

Y'all are approaching this the wrong way

Warnings ain't gonna work

Also gtg be back in 8 hours

You have no way of knowing if they've looked at the messages or not

@Lyxal They'll work more than if we didn't have them

10:56 PM

@cairdcoinheringaahing let me finish

No, warnings alone isn't enough

top quality banter incoming

@Lyxal No :P

You need to use the users webcam to track their eye positions

I'm expecting a shitty pun

is that it?

10:57 PM

Yes

oh

well that was very underwhelming

@Lyxal Nah, ask them to provide a sample golfed program :P

That's more anticlimactic because y'all interrupted

Smh

TBH, raising the asking privilege to 10 rep could work

You hack into their camera, and if they don't look at the warnings, their computer crashes

and if you say it can't be done it can

Even though I didn't read the article it still says it does

10:58 PM

lol

Play really loud noise at a pitch that only cats can hear, so that they get clawed to death

^

Or some other punishment

what if a cat tries to post a question?

that would be very offputting

Only if they don't look at the warnings

seems reasonable

10:59 PM

@rak1507 Cats can type better questions than half the new users, so give it an upvote and consider answering :P

Further, if they don't read the messages (verified by eye tracking software), you get their location from their ip address and send a Swat team to their house

Missile strike

The only way we're going to cut down on spam is with violence and invasion of privacy

Or by making it harder to do

A simple text box saying "oh look at me now dont post crap" ain't gonna cut it

No we need to get the message across with unnecessary force

11:01 PM

Raise the rep required to ask to like 2. Do literally anything on the site and you can ask

Nuke them until they ask good questions I say

Add a required course to every kindergarten program on how to use CGCC

Brainwash them young and no violence needed

@rak1507 is that more what you were expecting?

Or we can flash messages on the screen for a single frame every few seconds and subliminally message them into doing what we want

@RedwolfPrograms thats no fun at all

@RedwolfPrograms ooh now that has potential

11:02 PM

@Lyxal slightly

@RedwolfPrograms "if you write a bad question, you have to set your house on fire and buy a new computer as punishment"

Or "if you don't read the warnings, you'll never sleep again"

Anyway I thought y'all were supposed to be afk/gone

*looks around*

Seems they are gone

Good

Now I can type my funny numbers in peace without having to worry about the British ruining my reddit moment

69. 420.

Now laugh and give me silver

Bubbler

I have an idea: you can't dismiss the modal for 30 seconds. It should work because the 30-second in-game ads work for me

A: Sandbox for Proposed Challenges

11:19 PM

@Lyxal flags :P

New Sandboxed Posts

0

Hanabi playing bot Overview Hanabi is a cooperative card game with limited communication. It won the German "Spiel des Jahres" award in 2013. The game can be played by 2-5 players, each of which has a hand of 5 cards (4 cards for 4-5 players), which they can't see themselves, but the other player...