@PyGamer0 Relevant XKCD ;)

@PyGamer0 Also Relevant XKCD ;)

@DLosc What did you have for the Quipu factorial?

€:

> This article is about the mathematical concept. For the four-person anti-Nazi Resistance groups, see [other article]

NGL, I don't think I've ever confused a mathematical concept with anti-Nazi Resistance groups

@AaronMiller Haha! I hadn't seen that one before. It's a good reminder for us IT folks who can't understand why anyone would struggle with e.g. uploading a file...

@AaronMiller I didn't keep the solution I wrote, so I'm not sure. I think it was five or six threads, though.

@Bubbler [smiles mysteriously]

@AaronMiller Classic geochemists. Unlike programmers, who can accurately tell that the average person probably only knows about brainfuck, lisp and Malbolge

I never really got that XKCD

Usually I see the oppposite

People assuming common people have absolutely no knowledge of their field and explaining obvious things

Um

@DLosc I'm working on a language that is a bit like this, but it's a general golfing language, not for regex

@BrowncatPrograms Obvious to whom? You? Maybe you know more about their field than the general public does.

@BrowncatPrograms My dad tried explaining what the grant he was working on was about by starting with "So you know what cholangiocarcinoma is right? Well..." :P

@BrowncatPrograms (Which isn't necessarily a bad thing)

@Dudecoinheringaahing "carcinoma" - don't both "carci" and "-oma" mean cancer?

idk

He's the cancer scientist, not me :P

@Dudecoinheringaahing I'd guess it's a form of cancer having something to do with cholesterol in the bloodstream?

Or heart?

@DLosc bile duct

So close :P

Ah, chol-er not chol-esterol

@Dudecoinheringaahing answer: carcinoma starts in organ tissues; sarcoma starts in "supportive" tissues (bones, muscles); leukaemia starts in white blood cells; etc.

cancerresearchuk.org/what-is-cancer/how-cancer-starts/…

(still doesn't quite answer the original question though lol idk)

@Adám Haha! Is this where Jon Skeet's current profile pic comes from? That's hilarious. ^_^

> -50 User was removed

>:

@Adám You have Prolog. / You spend hours carefully crafting a magic spell that will declare the princess as rescued. / Prolog tells you, "no."

@AaronMiller I do recall that my factorial program worked for an input of 0, though. >;)

:(

@user When the Quipu interpreter takes input and there is no input, it returns null. Should it return 0 instead?

I don't think it should return 0--then there's no distinction between end-of-input and a 0. (Not that null is useful either.)

-1 maybe?

Maybe, yeah. I would prefer that Quipu could also handle negative integers as input, but that doesn't seem to be the case.

(Note "doc" in singular.)

what is HST?

High Speed Trading.

ah right

@Dudecoinheringaahing You jump to a thread that doesn't exist and fall to your death

The doc, for reference.

@Adám Ah, I am now fluent in K, very nice :P

tbf, what more do you really need :P

You could eliminate the "maybe" problem by destroying the universes where it doesn't work, so that the only universes that survive are the ones where it worked, and everybody lives happily ever after

But, per the nature of the infinite worlds theory, it will work in an infinite number of universes :P

@Dudecoinheringaahing happy now?

Not really, but that's got nothing to do with your message :P

:/

@Dudecoinheringaahing Are you OK?

Yeah, dw :P I'm not unhappy, but not really happy, if that makes sense :P

Yeah, as much as 0 is neither a positive or negative number.

Wait, the +-*/% knots don't merge with the previous ones?

No wonder this wasn't working

@Adám Indeed. I'm in a very zero mood today :P

frick, that's too hard to read

Time to think of the most psychologically troubling sentence you've ever read in your life :P

yeah lol

Y'all ready for this:

I regularly use PHP for fun, and see nothing wrong with any part of the language :P

Should I flag that as offensive?

@Dudecoinheringaahing you didn't put "lol" or ":P" at the end though! :P

@Dudecoinheringaahing Troubling, but not the most troubling thing ever :P

@AaronMiller All right, I've got a 68-byte factorial program.

Okay, so which sounds less complicated grayscale -> chromatic or chromatic -> grayscale? I personally believe the latter is less complicated, but, any second opinions?

@pxeger My bad :P

@Tacoタコス Depends if the former implies adding saturation (and hue) where there isn't any.

@Adám I'd say no IMHO

That sounds like it would make for a very complicated challenge

@AaronMiller Yes

@Tacoタコス In that case, the latter is much more complicated.

@Adám chromatic -> grayscale? Isn't that (in its simplest form) just an average? I know there's the weighted method too, but I didn't think it was complicated. Perhaps that boils down to the accuracy?

Sure, you could take a simple approach like that, but still, grayscale to chromatic is nothing more than taking 3 of each value.

Nice.

Qbasic really is the best language, though

I wrote really cool stuff in it.

I still say it's a better language for beginning programming than most of the languages beginning programming is taught in these days.

The "given its exact size" joke is based on an abortive plan of mine to implement tinylisp in QBasic. I realized that memory management using just fixed-size arrays was going to be a pain and scratched the project.

Just learnt that @RGS works at Dyalog - another code golf user

@Adám @rak1507 usually called hft ("frequency")

Well that was fun...our school has an early release day today, so lunch is a bit earlier than usual. I guess the lunch people didn't know that, so when we got there the cafeteria doors were shut. They're these thick steel garage-door-like things, and by the time they were opened like 200 people had queued up behind them. So there were hundreds of hungry high school students ducking and crawling underneath the doors as they opened, and swarming everywhere there was food :p

@Adám @Dudecoinheringaahing that's a reference, not doc

@ngn Ah, right. My bad.

@ngn reference ⊂ doc

@Adám john s's version: k sets things up so the princess doesn't need rescuing

@pxeger Also Karta, but I don't know what his user name is here.

@pxeger right, a proper subset

@ngn (I couldn't find the Unicode symbol for a not-necessarily-proper subset)

@ngn Sure, but it doesn't have that defeatist vibe.

@pxeger ⊆

@Adám It couldn't find it easily copy-and-pasteable on Wikipedia ;)

@pxeger but k has docs (not the single-page ref)

@Adám ? I'm not sure I follow; if the simplistic form of chromatic to grayscale is the average, then wouldn't you need to increase each component by a factor of 3 (may have phrased that wrong, so r * 3?)?

> not-necessarily

@Tacoタコス Depends on the representation. I was thinking converting an 1×y×x array into a 3×y×x array.

@pxeger speading uncertainty and doubt? :)

@Adám still not tracking in all honesty

@ngn no, because I know that I know nothing about K, so I didn't want to give any false info

@Tacoタコス for an image of resolution 2x2, you store a 2x2 array of pixels. For greyscale, each pixel is just a value from black to white. For colour, each pixel is three values representing amounts of Red, Green, and Blue

@Tacoタコス If I had an image represented as a 3D array of [0,255] values, 1 layer (luminosity) and then rows and columns, and I wanted to convert it to RGB, I'd make that 1 layer into 3 identical layer (luminosity of R, G, B).

So to convert colour -> greyscale, the easiest way is to take an average of the three R G B values and treat it as an brightness value for some shade of grey/black/white

@DLosc It took some doing, but I managed to get a 64 byte factorial program in Quipu that works for n=0 :)

Very nice!

@pxeger Although that is very wrong.

?

@pxeger @Adám is right in that it's not accurate, but it's the simplest way to do it, from my understanding

For greyscale -> colour, the simplest way is to replicate the brightness value three times, once for each of red, green, and blue, which still produces an image that looks grey (because the R G B channels are all equal), but is encoded as an RGB image

#FFFF00 is about the brightest colour you could get, but it'd translate to #AAAAAA

So my confusion with the array issue earlier is that I was looking at this from a single pixel standpoint instead of an image.

And searing #00FF00 would become a very dark #555555

ah yes, you'd need some adjustment for the eye's different relative brightness perceptions of red, green, and blue

Even without that, a better conversion would be to use max(R,G,B) rather than avg(R,G,B)

@pxeger The weighted system takes this into account, but still isn't the most accurate from my understanding; it simply improves the result.

@DLosc Do you want me to post it, or do you want to try and find it yourself?

Go ahead

So I suppose to run the challenge by everyone before I flesh it out in the sandbox; given a color represented by a six digit hexadecimal value, output a six digit hexadecimal value, representing that color in grayscale.

CMC: (foo :-) or (foo :-))?

vtc as unclear

@Tacoタコス sounds like not that interesting, the answers will just be "split into 3, convert from hex, average, convert to hex, join"

Grrrr, back to the drawing board then lol

Thanks for the feedback!

@DLosc I just remembered there's a factorial challenge, so I'll post it there

@pxeger (foo :-)) is the obvious choice

> The APL Orchard chatroom, where a small number of brilliant people convened to discuss all things APL

Why anyone would even consider (foo :-) is beyond me

You wouldn't write (x + (-4) and consider that to have balanced, nice looking parentheses

(foo :-) just looks like foo :- in parentheses

^

TBH I think avg is closer

Yeah, I realise that. But it is because of the weighting. We see the blue as darker.

@DLosc posted

Ah, you special-cased 0. Interesting that my approach that doesn't special-case the input was longer.

I guess it's because a shorter leftmost thread doesn't necessarily save bytes.

@DLosc Now I'm curious; how'd you do it?

\/1&1&1&2&
  ++[][][]
    **0&/\
    3&[]
    >>--
    1&1&
      <<

Wait

No, don't wait

I thought I saw an obvious golf, but...

Wow, yours looks so much cleaner than mine lol

a) it actually adds 3 bytes and b) it doesn't actually work :p

lol

@BrowncatPrograms this is informal prose, remember...

Formalness of the prose doesn't change anything, right?

@DLosc Here, the avg is way too dark compared to the background.

Of course, the max seems overlit, but the contrast matches the orig.

I'd say they both have problems. To me, the colored pepper still looks darker than the background, so the max is very wrong; but yes, the avg is too dark.

@Adám what does the harmonic mean look like?

Here is an average of avg and max:

@pxeger Not sure how to compute that for RGB values.

@BrowncatPrograms GRT keeps opening empty FQ and FA tabs for me

Seems to happen only when there's an item to review as well

Weird. I don't think I can repro that.

I'll have to wait though

@Dudecoinheringaahing I had that happen once; if I kept reloading the actual review task appeared after a minute or so

probably caching

It's weird, it just opens the https://codegolf.stackexchange.com/review/first-questions page, rather than to a specific review task

Think it happened with Suggested Edits once as well

This is an unlikely possiblity, but did you disable the userscript at any point after first loading the GRT tab?

^^ is weighted according to CIE 1931

@BrowncatPrograms Nope, GRT's always been active

I'll look into it later, there might be an issue with the foreground being desynced with the page being loaded in the background

^ CIE 1931 of the first image.

@Adám Try these images:

@Dudecoinheringaahing Not much difference ^ imo.

@AaronMiller Trying a new factorial approach. Finished writing the code and hit Run. Input, 4; output, -11. X^D

oh, so close :P

Well, it helps if I remember that /\ is output and \/ is input. Can't count how many times I've mixed those up. Now it's outputting 1 instead. :P

Aha!

Okay, I've got a 50-byte solution. ^_^

O.o

Hint: it uses only three threads.

Did you special-case the 0 input?

Not as such. But I did add 1 to the input.

Alright, let's see if I can figure it out :)

My fancy JS brainfuck interpreter is 124 lines long, I feel like that's too many

Wasn't the first BF interpreter 240bytes?

Compiled, probably

Okay mine's actually only 96 sloc

And to be fair I do waste 18 of them on fancy I/O and custom error types :p

Fancy != golfy, so I don't see why that's a problem

Yeah, but for such a minimal language it feels like I must be doing something wrong :p

CMQ: do y'all have ice cream trucks where you live

reason for asking: i just heard one outside - it's been a really long time since I last saw one as far as I remember and kinda forgot they were a thing

Not in my current neighborhood, but there were where I lived a few years ago

also, CMQ: do y'all have knife sharpening trucks where you live

No

I have never heard of that lol

@BrowncatPrograms ah. I don't think I've actually seen one where I currently live but I don't really remember; I think last time I saw one was where I used to live (moved like 3 years ago)

@BrowncatPrograms lol yeah i don't think they're nearly as common

i've only seen one like once or maybe twice in my life in my old neighborhood

I don't think there's really any organization of ice cream trucks, it's probably just people buying trucks and going around selling ice cream on their own

So I wouldn't really expect any geographic patterns

yeah, i don't think there's some organization to it. which if i'm being cynical sounds really prone to use by people who want to hurt people / specifically children

now that i think about it you probably need to register as a food business (+ there's probably extra for food truck businesses specifically)

@hyper-neutrino No, but you can take your knives to any half decent butcher and they'll sharpen them for you here

I don't think there's any butchers where I am, you just go to the store and there is mysteriously meat there

@hyper-neutrino google says this is almost entirely exclusive to toronto