The Nineteenth Byte - 2017-03-02 (page 3 of 3)

« first day (2224 days earlier) ← previous day next day → last day (2626 days later) »

10:08 PM

@Riker Yep it's awesome, I think we have a feed into Primes&Squares.

@LliwTelracs It doesn't seem all that hard. For every pair of stations generate a list of all paths between them (depth first search would work). Then for every order of station visits (e.g. A->B->C, A->C->B, B->C->A...) test every combination of paths until you find one that doesn't overlap.

Destructible Watermelon

Idea for a challenge: given a dictionary of string keys and string values, where the values are links, and one string, output whether, by following the link of a string key, by moving onto the string value, if it is a key in the dictionary, all the values eventually reach the one string

like how "every" article on wikipedia eventually leads to philosophy

@Riley yeah, that would take ages

@JanDvorak At least I have already reduced the map to the valid rails only

@JanDvorak It would get the job done eventually

10:22 PM

About the evolution of powers of two. I extracted the codes. Edit distance of #18 is 12 and that of #44, #47, #52, #56, #60 and #66 is 11.

Only took me about 5k characters of conditional statements

Proves, I guess, that editing and calculating edit distance by hand is too error prone :P

/me Beats Sami with a wooden sign that says the single word "grandfathered"

s/56/57/

(Copying numbers by hand is hard too)

@Riley The adjacency matrix for the rails is something like a 70*n by 70*n table

10:38 PM

wow this illusion is way better than the checkerboard

user image

the spot in the middle is the exact same colour

the same pixel value, or the same texture color?

Comparing the pixel values in face of wildly different lighting conditions would be a fatal mistake - an yet that's what everyone keeps doing when they call it an illusion

@SamiLiedes I checked #18, And the two unaccounted for differences are two homoglyphs that don't actually effect the challenge. Seems like the mistake of a text editor more than anything.

user image

@JanDvorak the lighting conditions are the same

those are the conditions of whatever room you happen to be in

@orlp which spot?

the middle of the x's?

@Maltysen yes

10:45 PM

Yeah, but those yellow and blue rectangles are either stained glass, or stained glass with very specific cutout in the middle, tuned to the camera position

Destructible Watermelon

oooooh

@JanDvorak what are you talking about

I could print this image out on paper

then you could take scissors

and cut out the middle x parts

Then maybe. But I care about what the image represents, not what images are used to represent the images.

bring them together, and see that they are indeed the same color

Your argumentation might pass here because you're vague enough, but it falls apart in the cylinder-on-checkerboard image

10:47 PM

@JanDvorak that's an unreasonable position, would this optical illusion only become real for you as soon as someone stitches it in real life? or paints it on a painting

the fact is that your brain interprets a color wrongly based on surrounding information

that's the illusion

Destructible Watermelon

yeah ok stop whining. the entire point is that you can't imagine them being the same shade in the checkerboard cylinder illusion

regardless of whether its because of the lighting

Then why does everyone keep calling the areas A and B in the checkerboard image "squares"? They clearly aren't. They're parallelopipeds.

Destructible Watermelon

what?

Perhaps they are and the image was simply stretched?

@JanDvorak I'm not interested in arguing against your strawman.

10:49 PM

I'm not interested in argueing.

I'm interested in code golf.

@ATaco: Also, in #66 one extra change is merely a change in whitespace at the end of lines, which I don't think probably matters.

I'm just saying your claim is being vague - and the less useful interpretation in real life.

And, the claims about the checkerboard image are blatantly false

CMC: The exponential factorial: Given an integer n, output (((...(1^2)^3)^4)...^N)

I take it you disagree with those other claims about "the squares being the same colors"?

I want you to delete that image

Destructible Watermelon

why?

10:53 PM

It offends me mildly - and your refusal to remove it offends me non-mildly

I don't think he can delete it anymore.

It's not an offensive image, any more than your saying someone's claim is vague and not useful in real life is offensive

Needs a mod or a bunch of flags.

Then trash please

Somebody cleared the "bunch of flags", unfortunately

more than one person.

10:54 PM

@Rainbolt I'm trying to be objective. Did I come up as ad hominem?

Martin's avatar moved to the front. Is that an unintended giveaway that he took mod actions?

@Pavel Isn't that always 1?

Yes, pretend it starts at 2

@Rainbolt basically, yes.

I accept suggestions how to reword what I said

10:55 PM

@Pavel that grows way too fast

Is point

make it more interesting by doing it modulo a second input :)

Sure, do that. It's a CMC after all.

oh actually it's fine

lambda n:1

that's my answer :)

1^2^3 = 1

Destructible Watermelon

@orlp yeah, I saw one like that before, with yellow and brown, but this one is better

10:56 PM

@Pavel (((..(2^1)^2)^3)..^N)?

1^2^3^4^5 = 1

1 <- Pyth implementation

I meant to say ((2^3)^4)^n

Sorry

Destructible Watermelon

that reminds me of when someone made a CMC that given input n output if n squared was prime

(((((A^1)^2)^3)^4)...^B)?

Let's just promise to not talk about optical illusions (air-quoted or not) here.

11:00 PM

@JanDvorak how about no?

@Pavel surprisingly fun to Golf in Mathematica: ##2^1&@@Range@#&

I've never had issues posting them before

a lot of people find them (mildly) interesting and discuss them

@MartinEnder Which version is that?

There's like 4 now

?

oh

your latest one

Ok

11:01 PM

actually no

I was about to merge the suggestions into:

it's 2^3^...^N

Revised CMC: Given integers A, B, and C, output ((...((((A^1)^2)^3)^4)^...)^B)%C

so wrong associativity

@MartinEnder I'm somewhat surprised to realize that I know how that works mostly, but what is ##?

11:03 PM

## is a sequence (like a splat) of all arguments. ##2 is the same thing but starting from the second argument. ##2^1 is syntactic sugar for Power[##2,1] then ##2 gets expanded so we get Power[2,3,...,N,1]

but like I said, unfortunately that's 2^(3^(...^N...))

@Pavel 05AB1E: ¹²GNm}³%

btw the challenge is equivalent to a^(b!)%c

Not familiar with what a splat is... is it just a set?

@orlp oh, really? huh

Bah, math is hard

@Pavel assuming positive numbers, (a^b)^c = a^(bc)

The fifth panel also applies to postmodernists.

11:06 PM

@Pavel no, it's just what the same thing is called in some other languages... it basically means that you insert a list without any "wrapper". {1,##,3}&[1,2] is {1,1,2,3} as opposed to {1,{1,2},3}, say.

@JanDvorak why are you so hostile? have I done something to upset you?

@MartinEnder So, it's a list that flattens into it's parent list?

@Pavel it's quite intuitive

let's say you have (a^3)^5

@Pavel basically

you can see that like this:

11:08 PM

@orlp No I know

a*a*a * a*a*a * a*a*a * a*a*a * a*a*a

I just forgot

3 groups of 5

but that's the same as 5 groups of 3

It's been like a year since I've done this whole Algebra 2 thing

@orlp Not you specifically, but I don't like that people keep saying "ha! the brain is so stupid for thinking this image represents a 3d scene, and confuses the 3d scene with the image that represents it", while they themselves make the same mistakes.

11:08 PM

or 1 group of 15

@JanDvorak again with the strawmen, I (nor did anyone else here) never mentioned that the brain was stupid for doing so, just interesting how the brain makes these kinds of mistakes

that's the whole point of optical illusions

I keep saying it's not a mistake

It's more like an assumption

and regardless of that, that's not the hostility I refer to

Can some mod trash this entire optical illusion discussion? Pretty please?

Well, I started with a question, to which you replied "what lighting condition are you talking about" as if the image was disconnected from any meaning

11:11 PM

the part that's hostile is using comics that, when used as a reply to me, would refer to me as being 'smug', asking other users to self-censor mild jokes, abusing flags for that, constantly putting up strawmen and then attempting to argue against those straw men, and just in general being very argumentative and hostile

Destructible Watermelon

I got a pack of optical illusion cards once. its pretty cool

I just really don't appreciate it

A targetted joke is never percieved as mild

and I, particularly, have trouble distinguishing mild/non-mild jokes. I can apologise for that.

I didn't mean to call you smug, what I meant to say that a big part of these illusions is that the claims are poorly defined.

Destructible Watermelon

basically its just that even after being told its essentially impossible to see the colours as being the same

without isolating them

I have a KOTH idea:
Players submit a program in a 2d, self-modifying language (possibly the same one for all contestants, haven't decided) That fits in a certain size bounding box
The programs are put onto the same toroidal grid at different spots, and they compete to see how long their programs can run.
The key is that they interact with eachother, so if a program puts a "terminate" program" command somewhere, and another runs into it, the program that ran into it ends.

Destructible Watermelon

11:16 PM

@MistahFiggins soooo 2d core war?

Now that's an interesting note - and one worth making - that the brain does not provide the low-level information. I suspect it's because this data isn't even present because the eye only detects changes, not constant values.

Essentially

Destructible Watermelon

impossible colours are also fun

I had a 500-character reply prepared, but then I decided that a comic reference would be better percieved than anything I could say

@JanDvorak no, the low level information is definitely there

11:19 PM

Esolang idea: The program flow follows an Hilbert curve

@orlp reference?

5

Q: How does the eye perform the equivalent of white balance?

Possible Duplicate: How can knowledge of human perception of color be used in photography? How does the eye perform the equivalent of white balance? Why is the camera not able to replicate it?

white-balance human-vision-system

That answer essentially says that color processing is intrinsically local, doesn't it?

> uses separate white balance on different areas

pretty much what the retinex filter emulates

- and has the same benefits as the retinex filter - it boosts contrast in favor of absolute brightness. It's a feature.

Destructible Watermelon

also, ever heard of eigengrau?

Eigengrau (German: "intrinsic gray", lit. "own gray"; pronounced [ˈʔaɪ̯gn̩ˌgʁaʊ̯]), also called Eigenlicht ("own light"), dark light, or brain gray, is the uniform dark gray background that many people report seeing in the absence of light. The term Eigenlicht dates back to the nineteenth century, but has rarely been used in recent scientific publications. Nowadays, the phenomenon is more commonly referred to as "visual noise" or "background adaptation". Eigengrau is perceived as lighter than a black object in normal lighting conditions, because contrast is more important to the visual system than...

The information that's lost is exactly the kind of information that you only really care about while trying to find the deficiencies of human vision

11:26 PM

@JanDvorak it says it happens in the visual cortex, which means the low level information is there at the eye nerves

The eye nerves carry differences, not (or only weakly) absolute values

at least in the peripheal vision

@JanDvorak you have to be more precise here

differences between what?

The change of brightness over time. The eye wiggles a bit to convert spatial edges into temporal edges.

@JanDvorak ah but that's the crux, that happens only per sensor in the eye

and each of the tristimulus sensors are doing that on their own

no 'white balance' happens

Except that the sensors seeing the flat areas stay completely silent. Only those that can see edges report any signal.

11:29 PM

you can not have white balance until you combine tristimulus values not only cross-channel, but also non-local

I'm not entirely sure about the details, unfortunately

@JanDvorak your theory also doesn't work for slow changing gradients, your eyes would not be able to perceive a very slight gradient in a direction orthogonal to your eyes wiggle

Try fixating your eyes on a very slight gradient and you will notice it starts to disappear

` en.wikipedia.org/wiki/Filling-in

@JanDvorak can confirm. Does dissapesr

@JanDvorak alright, but here's the crux

you claim the whitebalance correction occurs due to this effect, but we also see that this effect works slowly on a gradient, then why does it work instantly on the optical illusion? even if the optical illusion is only flashed for a fraction of a second?

it's a different mechanic

11:36 PM

I expect most of it is given by the brain reconstructing a 3d scene from the image

if you see strawberries under strong blue light, you don't want your brain to think they're mold-grey and should be thrown out.

the brain is reconstructing a lighting scenario, and interpreting the image under that assumed lighting scenario, but that is not something that happens at the low level

that only happens in the visual cortex

the (objectively correct) low level information is there, and does make it to the brain

however we cannot access it, even if we are told of what is happening

What happens if you flash the checkerboard image while focusing on trying to get the raw data?

@JanDvorak I don't know, or even how to test that

but I do have another testable idea

you can make many different versions of the optical illusion I showed

On one hand, the raw stimulus should be more accurate, on the other you have trouble focusing the right pixels in the visual field

with different discolorations etc

and show people the image for 0.5 sec

and then ask them 'are the two centre colors the same, or not'

and then see if people do better than random chance

(assuming you told them the optical illusion beforehand, and that they must try to ignore that)

11:41 PM

Yeah, be sure to communicate clearly

my hypothesis is that people will not be able to do better than random chance

@ConorO'Brien Not that I'm aware of

What color difference do you want them to discern as different?

Also yay i got an answer with >10 votes

some colors are difficult to tell apart even when presented with the exact same surroundings

The same kind of difference as is between the "white squares" (lit) and the "dark squares" (lit) on the checkerboard picture?

Do let me know the results

11:44 PM

I'm not planning on executing the experiment, sorry

d'awww :-(

@0 ' Congrats for doing 3 meta-reviews today.

@WheatWizard I nearly doubled my total number of meta close vote reviews today

One nice effect I like is the color afterimage effect. It doesn't work for me very well, though.

I'm about to post this soon. Does anyone have thoughts on it?

2

A: Sandbox for Proposed Challenges

Happy birthday V! code-golf ascii-art string V is a programming language that I wrote so that I could use and extend vim for code-golf challenges. The very first commit was on March 3rd, 2016, meaning that today V turns one year old! Woo-hoo Over V's first year of existence, there have been ...

11:58 PM

Looks good to me.

« first day (2224 days earlier) ← previous day next day → last day (2626 days later) »

Transcript for

Mar '172

The Nineteenth Byte

The Nineteenth Byte: General discussion for codegolf.stackexc...

join this room
about this room

00:00

06:00

12:00

18:00

all times are UTC

site design / logo © 2024 Stack Exchange Inc; legal

mobile