Cubically

TehPers

01:34

@MDXF Oh thanks

I just deleted my answer on another question. Thanks for letting me know xD

MD XF

04:21

Oh lol... I had to do that too a while back

Feeds

16:09

MD XF has made a change to the feeds posted into this room

MD XF

16:25

→ 5 messages moved to trash

TehPers

18:03

@MDXF How does the interpreter handle U³2²4?

Does it rotate layer 3 twice, then layer 2 four times?

Also, which layer does U³²3 rotate?

Also, do you have a table containing the values of each character? I'd assume the first 128 characters are the same, but where do the superscripts fall in?

For now, I'll just assume the superscripts are decimal 128-137, but I'll change that once you reply.

MD XF

18:30

@TehPers No, superscripts are command-dependent, not turn-dependent. So U³2²4 is equivalent to U³2 U³²4.

@TehPers It tries to rotate the 32nd layer, but if the cube is not large enough to handle it, it doesn't do anything.

@TehPers Cubically's codepage (table) is printable-ASCII-compatible, but the control codes are different (with the exception of the newline and the tab). The superscripts are 0x00-0x09 in Cubically's codepage and seemingly random values in UTF-8.

TehPers

Okay, thanks!

I'm working on implementing the codepage

MD XF

18:49

area51.stackexchange.com/proposals/113168/…

TehPers

@MDXF I think it'd be better if this behaved as "rotate index 3 twice, rotate index 2 4 times". It just makes more sense in my head. I can't really justify it past that xD

@MDXF ?

MD XF

@TehPers Proposition for a Rubik's Cube Stack Exchange :D

TehPers

@MDXF I think that'd be better off as part of the Mathematics SE site

MD XF

A lot of cube-related things aren't particularly on-topic there

TehPers

It isn't?

I feel like it should be o_O

If not there, then where?

MD XF

18:52

The first two things in Rubik's SE's description would - mathematics, theory - but the other things are less so - algorithms, solutions, permutations, mechanics, or design.

Of course, you might be right, it might get closed as a dupe of that; however, PPCG's "Esolangs SE" proposal has gotten closed as a dupe of Programmers SE... and questions about esolang development are most definitely not going to be well-received there.

TehPers

Idk

I'm gonna look at how Jelly handles the codepage

MD XF

19:53

Jelly performs conversions, the C interpreter doesn't. Idk which would be easier to do in Lua

TehPers

I already have utf8 to char array

MD XF

Oh sweet

TehPers

Basically, I'm converting the string (program) to the UTF-8 character array, but each element in the array is actually the byte value of your codepage, or -1 if it doesn't exist in there

Then it can be converted to characters, where � is used for unknown characters

� is not part of the codepage. It's just used for characters without a spot in the codepage, like decimal value 184, or decimal value -29

MD XF

Good idea

TehPers

20:09

@MDXF It's just for the way my conversions work, not a suggestion for the language xD

Anyway, will superscripts only affect layer selection?

MD XF

I know lol, I'm just saying it seems like a simple, efficient and foolproof way of doing it

TehPers

Or will they be used for things in the future?

MD XF

Superscript numbers will only be used for layer selection

I haven't decided whether or not superscript letters/symbols will be included in the codepage

TehPers

@MDXF How else will you handle them?

MD XF

That's the thing; I don't know. They will likely be useful for something in the future; they're not right now.

TehPers

20:13

@MDXF Why not use subscripts instead?

MD XF

How come, i.e. what makes subscripts better than superscripts?

Or do you mean use them for something else?

TehPers

@MDXF Superscripts are used for exponentiation, and subscripts are more used for indexes in math

Like t_0, for example

MD XF

That is true

TehPers

Or F_0, F_1, F_2, etc

MD XF

Has anyone used the superscripts on PPCG yet?

TehPers

20:15

TIO doesn't support them yet

So no

MD XF

Really??

tio.run/##Sy5NykxOzMmp/P8/6NBOIyuD//8B

TehPers

I tried earlier...

Oh huh, must have been recently updated?

MD XF

They're a bit broken but AFAIK they've generally worked for a few days

TehPers

shrug

Regardless, nobody's used them yet

Just did a search

MD XF

On an unrelated note, if you install chatcommands (via tampermonkey or something), typing /shrug in chat will auto-insert the shrugging emoticon xD

@TehPers Oh ok that's good

TehPers

20:18

@MDXF shrug

MD XF

Just in case someone had plans, until we find some other use for superscript numbers, I'll make them compatible, so you can use superscripts or subscripts to define layer depth

@TehPers lol

TehPers

@MDXF Ok, I'll only support subscripts

MD XF

Sounds good

Oh phew, Unicode subscripts are actually convenient

TehPers

@MDXF github.com/Cubically/cubically-lua/blob/codepage/…

TehPers

21:07

@MDXF In response to the issue with arguments being passed in from input, maybe allow superscript x do that? The thing is, this can also be done by setting the notepad to your input (or whatever) and decrementing it in a loop

What is the exec command?

MD XF

@TehPers That's a great idea!

@TehPers Which exec command?

TehPers

@MDXF Nevermind

Also, check the issues on your repo

@MDXF What would be used for input, what would be used for face value, and what would be used for notepad?

If you need uses for superscripts, you could do that I suppose

I'd have preferred a bit of consistency, where digits are constant values and superscripts are cube/notepad/input values

But it's too late for that

Hmm

I have another idea.

MD XF

I'm a bit confused

TehPers

Um

So you want to rotate the face 2 times? You can do R2.

MD XF

Yeah

TehPers

21:21

You want to rotate the face a number of times equal to the input, you can do R2 with a dot over the 2 or something

or hm

Idk

MD XF

Why 2?

TehPers

I want this to work in subscripts too

2 was just a random number

Give me a sec to think

MD XF

Meaning you want R₁(foo) to work, where (foo) is whatever takes input as the argument?

TehPers

2ͯ

That'd be the value of face 2

or something like that at least

MD XF

Oh I see!

TehPers

21:23

Basically, it'd be cool if you could do Rx, where x is the value of a face

Also Rₓ, aka rotate layer x

MD XF

So R2 would rotate the right face twice, while R2ͯ would rotate the face index stored in face 2 twice

TehPers

It'd rotate the right face x times, where x is the value of face 2

MD XF

Yeah exactly, I just couldn't think of a good way to phrase it

TehPers

And Rₓ₂3 would rotate the right face, x layers in, 3 times, where x is the value of face 2

Actually

Rather than messing with the numbers

Why not create another command?

Use a R with a circle over it or something

MD XF

Hmm

TehPers

21:30

The thing is, you'd need to support passing variables in as arguments and subscripts

If you're only going to have face rotation as letters, then you might as well create variants of it

the issue is, then they'd take 24 characters from the codepage on their own

Versus using superscripts for constants, and the digits for the value of that face (it's too late to do this though)

MD XF

Well, I do have at least 127 more codepage slots open.

TehPers

That's true

ǲ is one character..?

Dz ǲ

shrug

MD XF

There are lots of much weirder Unicode characters, trust me xD

TehPers

Number Forms

Number Forms is a Unicode block containing characters that have specific meaning as numbers, but are constructed from other characters. They consist primarily of vulgar fractions and Roman numerals. In addition to the characters in the Number Forms block, three fractions were inherited from ISO-8859-1, which was incorporated whole as the Latin-1 supplement block. == List of characters == == Block == == History == The following Unicode-related documents record the purpose and process of defining specific characters in the Number Forms block: == See also == Unicode symbols Latin scrip...

You could try using the roman numerals xD

MD XF

@TehPers which Unicode character is this?

TehPers

21:37

@MDXF ͯ

MD XF

I see an empty message...

Is that a really really really tiny 2?

TehPers

@MDXF No lol

Unicode subscripts and superscripts

Unicode has subscripted and superscripted versions of a number of characters including a full set of Arabic numerals. These characters allow any polynomial, chemical and certain other equations to be represented in plain text without using any form of markup like HTML or TeX. The World Wide Web Consortium and the Unicode Consortium have made recommendations on the choice between using markup and using superscript and subscript characters: "When used in mathematical context (MathML) it is recommended to consistently use style markup for superscripts and subscripts...However, when super and sub-scripts...

overscript miniscule x

xͯxͯxͯxͯxͯ

MD XF

Oh lol. I meant which character is 2ͯ

TehPers

r᷊ͬr᷊ͬr᷊ͬr᷊ͬr᷊ͬ

2ͯ`` is two characters lol (I give up)

MD XF

Ah

We could use the U+1D7CE through U+1D7D7, mathematical bold characters

Or the U+1D7F6 monospace numbers

TehPers

21:43

Are those digits?

Sure, that'd work too I guess

MD XF

I think everything around U+1D7F6 is digits in different fonts

Scroll down to around the bottom of this

TehPers

I like the bold ones since they stand out

But maybe the double striked ones?

𝟞

MD XF

Either bold or double-striked, monospace will look the same as normal characters in most text editors and IDEs

TehPers

They look very different, easy to spot the difference

MD XF

Yeah. I don't really care, which would you prefer?

TehPers

21:46

➀ lol

I prefer the double striked ones

graphemica.com/%F0%9D%9F%B7

More options on that page

MD XF

Assuming this is the method used anyway lol, I've been adding all (two of) the possible methods here

I like it though

dingbat negative circled 1

What even is Unicode these days

TehPers

Dingbat, that's what.

MD XF

Lol

Oh my it goes farther: ➊ dingbat negative circled sans-serif digit one

TehPers

Yep.

MD XF

Is there anyway to rotate layer x? well with all these insane Unicodes there's probably U+FOOBAR subscript double-struck wingdingbat 2

TehPers

21:51

Lol.

MD XF

Assuming I understood your comment correctly - you mean is there a way to pass an argument to R (or L or whatever) telling it to turn the Xth layer, where X is the value stored in the argument passed?

TehPers

Yes.

"Rotate layer x from right face twice, where x is the value of face 2" for example

MD XF

Ok... that's also a good idea. Maybe that's what the bolded numbers could be, if we use the double-struck numbers for the number of turns?

So R𝟏𝟚 would rotate x layers in from the right face y times, where x is the faceval of 1 and y is the faceval of 2

TehPers

It doesn't look consistent though

I'm thinking it'd be better to just ditch all of this and use a fancy looking x or n to indicate you want the input from the arguments

Though again, I'd prefer to just rewrite the language from scratch

Actually, hm

Would it be a bad idea to make a new RLUDFB command using a unicode character

And the layer will be taken from the face value instead?

MD XF

It would be easier than rewriting the language and wouldn't invalidate all Cubically answers xD

I'm a bit confused about what's inconsistent about R𝟏𝟚; I might not be understanding what you mean by "consistent"

TehPers

21:57

Subscript selects the layer

MD XF

Oh yeah

Well now I'm just wondering... how useful will it actually be to take the layer as input?

It's something I'd like to add, and a good feature, but is it useful enough that it's worth either rewriting the language or using up most of the codepage slots?

TehPers

Being able to store a value on a layer would be cool...

Select layer x, rotate it to face 0, do something, rotate it back

I don't know how possible this is

Hmm

MD XF

I have a (possibly) better idea (might be alot worse)

I could make the notepad a tape instead of just a single value

TehPers

Anything that lets us store more than one value at a time would probably make this language Turing complete

MD XF

It'd make Cubically alot easier to use, but at the same time, it would remove some of the fun and the story behind it

Yeah actually it would completely remove the necessity to use the cube at all... it would basically make the language an extension of Brainf**k, especially since Kamil found a way to increment/decrement the notepad regardless of face values

TehPers

22:04

@MDXF I don't really know what to do about that.

Just being able to select a layer would probably be enough

R₁1U₁1L₁1

MD XF

Yeah maybe...

TehPers

That will set the value of the middle layer in a 3x3 to 1

R₁1U₁1L₁1D₁1 this does it cleanly

Man, I don't know, maybe you can use the subscripts to select the value of a layer of the face?

I can't think of a way to store multiple values on the cube

MD XF

Didn't Kamil find some 'cube incrementor'?

TehPers

Yeah, but you can't store 2 values on the face

And the max cube value is 1259 for a 3x3

MD XF

> multiple values

duh >.<

@TehPers How is that number so freakishly close to 1260?

TehPers

22:08

The max period is 1260

giving you the values 0-1259

MD XF

Oh so they're related. I see

TehPers

... on a 3x3, at least

MD XF

Yeah

TehPers

Regardless, you can't pull a value from the cube, manipulate it, and revert the cube

If you take the value from the cube, and you store it in your notepad, you can't return the cube to the state it was in without messing up your notepad value

Meaning you can't duplicate a value

MD XF

Speaking of 1260, I'm trying to find the max periods on larger sizes

1

Q: Determine the highest order of an element of a Rubik's Cube group

The period of a sequence of moves on a Rubik's Cube is the number of times it must be performed on a solved cube before the cube returns to its solved state. For example, a $90$° clockwise turn on the right face has a period of four; a $180$° clockwise turn on the right face and a $180$° turn on ...

TehPers

22:11

@MDXF How does that help you duplicate a value?

MD XF

Actually, never mind. It's useful for returning the cube to a previous state, just not an unsolved one. I was distracted sorry xD

Another question: Would X, Y, and Z commands be useful? They'd do what they do here

TehPers

@MDXF From what I read a while ago, this is really difficult to calculate

@MDXF I can't connect to that site

MD XF

That's odd... some sort of internet restriction?

TehPers

@MDXF I mean, I'm at work if that counts.

MD XF

Lol maybe. My work restricts the internet archive for some reason

TehPers

22:18

shrug

MD XF

Anyway. X is basically R1R₁1R₂1

TehPers

That'd do exactly nothing...

MD XF

Y is U1U₁1U₂1, Z is F1F₁1F₂1

@TehPers how so?

TehPers

@MDXF Because it'd be the equivalent of turning the cube sideways in your hand.

Pretty sure most cube solvers don't even rotate the middle face

MD XF

Well that's the point; it'll switch around some face indexes. Might be useful, might not, doesn't hurt

TehPers

22:24

I guess so...

I'm still looking at how to store multiple values at once

A lot of time seems to be spent on trying to do constant operations, like decrement and increment

It'd be neat to use the cube to store memory, but you also need it to provide constant values

Hmm

I don't know.

I'm gonna make a spinoff of the language with my ideas.

TehPers

23:05

@MDXF I'm rewriting the language and reorganizing the codepage, as well as changing a lot of operators. I'll show you when I'm done, and hopefully that gives you some ideas.

MD XF

23:21

All right, sounds good

I'll look into redesigning the codepage

Maybe, since we don't have any use for superscript numbers, subscript x could set the layer to x, and superscript x could set the layer to the face value of x?

It works, it's simple, and mostly consistent.

TehPers

@MDXF You'll see what I have in mind.

github.com/Cubically/cubically-lua/tree/variant

I gotta go, but that's the start of it

MD XF

So pretty similar to what I said earlier? since subscript for constant layer selection and superscript for face value layer selection

Transcript for