The Nineteenth Byte - 2018-04-03 (page 1 of 3)

« first day (2621 days earlier) ← previous day next day → last day (2212 days later) »

Esolanging Fruit

12:07 AM

I don't really like Google Docs for any document that's intended to share with everyone on the internet, since my Google account has my real name and I'd prefer for that not to be linked to the doc. In fact, Google made it recently possible for editors to see the view history of a document—a feature which I immediately disabled, but which still shows that Google isn't particularly interested in preserving privacy.
Not particularly surprising considering Google's business model is literally selling your data.

New Sandboxed Posts

0

A: Sandbox for Proposed Challenges

ASCII Line Extender code-golf ascii-art Introduction I have an ASCII grid of size 10x10 +--------+ | | | | | | | | | | | | | | | | +--------+ Occasionally, slashes and Xs pop up, but they're too small for me to see! +--------+ | ...

Ｕｎｉｃｏｄｅ-ｏｎｌｙ

@EsolangingFruit ... what.

Esolanging Fruit

Care to expand?

Ｕｎｉｃｏｄｅ-ｏｎｌｙ

if you want to share something with other people and want them to edit i'd say those are things you want enabled by default

Esolanging Fruit

No, but it's read-only anyway.

Ｕｎｉｃｏｄｅ-ｏｎｌｙ

12:10 AM

> editors

if they're not editors then they can't see view history

Esolanging Fruit

That includes the creator.

Ｕｎｉｃｏｄｅ-ｏｎｌｙ

well yeah? some people would need that functionality - it's better to have it enabled by default for those people that have no idea how to use a computer than have too many people feature-request it because they don't know it's a thing

12:24 AM

Congratulations to @DJMcMayhem for becoming our first Steward badge!

4

Ｕｎｉｃｏｄｅ-ｏｎｌｙ

holy what

Review badge?

codegolf.stackexchange.com/questions/158371/… /

@Downsheep up for team defense fort 2?

God I hate stats/

oh come on guys, no stars for DJ?

that's a pretty damn impressive achievmenet

12:27 AM

Currently invalid answer.

Got bored, tried to do a regression model for (x,y,z) => w by hand. So many fractions

This loophole was posted on Apr 16, and

Anonymous

12

A: How should we handle answers that predate standard loopholes?

Leave alone if they were funny once, delete if they were not We have a few loopholes that are actually common flaws in the spec, and exploiting them in the appropriate challenges could be considered clever. For example, answers that violate Abusing native number types to trivialize a problem ...

+0/-0 and +1/-0. Not sure if that qualifies as funny.

Anonymous

Essentially, that answer boils down to "we don't delete answers that were once reasonable and valid, but the community decided to ban something they used"

Anonymous

12:36 AM

It might be worth noting that I don't agree with that policy. But, community consensus has established it as the policy we apply, so I accept it and follow it.

@Picard oh crap lol just restarted computer

@DJMcMayhem it even links to Hoverboard ... that's when you know it's a huge network of stuff.

@Downsheep lol, in medieval rn

should be over quick

You still have the 4-1 names?!

@Picard ty

12:41 AM

np

@Downsheep done with game, you up?

@Ｕｎｉｃｏｄｅ-ｏｎｌｙ It wasn't on the star board when I posted that. Everyone sees a different number of starred posts based on their browser window size.

Ｕｎｉｃｏｄｅ-ｏｎｌｙ

@mbomb007 protip: click "show all" :P

@Zacharý Who? What name? ...

Downsheep, Picard, Unicode-Only (April 1st)

I think usernames can only be changed once every 30 days?

12 hours ago, by user202729

(Downgoat -> Downsheep, ASCII-only -> Unicode-only, quartata -> passata, Riker -> Picard)

You missed quartata.

Nope, quartata changed their name back.

12:53 AM

I'm planning on changing mine sooner or later

might keep it on ppcg for a while tho

So quartata's name changed has Pas sa t a ?

he's back to quar atm

Ｕｎｉｃｏｄｅ-ｏｎｌｙ

@user202729 yes

@Ｕｎｉｃｏｄｅ-ｏｎｌｙ idk if I pinged u already but can u look over the outline for VSL mem management: hackmd.io/dfa4KIwWRxu3RXGdsw_5kQ

Ｕｎｉｃｏｄｅ-ｏｎｌｙ

@Downsheep literal denotes unmanaged value?

sounds like not great idea

Anonymous

1:05 AM

@user202729 Usernames can only be changed once every 30 days. However, it's possible to have different usernames on different sites, and you can change your parent site for chat as often as you want.

@Ｕｎｉｃｏｄｅ-ｏｎｌｙ you are saying everytime I do 1 it should be reference counted?

that means there would be overhead for every int operation

since we have to wrap, unwrap, etc.

Destructible Lemon

@Ｕｎｉｃｏｄｅ-ｏｎｌｙ i don't 100% remember this

Ｕｎｉｃｏｄｅ-ｏｎｌｙ

@Downsheep no like string literal maybe? or dict/array literal

Why does searching for "Easter" giving messages containing "egg" ...

SE search is broken.

They're not even posted by NMP...

Destructible Lemon

its an easter egg

1:18 AM

LOLOLOL

mdr

Amazing

Destructible Lemon

also notice that the perma links don't work

permalinks to what?

Destructible Lemon

the messages

click on the links

and the messages you get change

this is an easter egg

/ april fools joke

Already tried, I thought they're just deleted.

Destructible Lemon

i reloaded it and got new messages

1:19 AM

Documented here.

Destructible Lemon

sucks to be you if you want to find easter related messages

... yes, I was looking for this.

Ｕｎｉｃｏｄｅ-ｏｎｌｙ

@DestructibleLemon no

Destructible Lemon

oh ok

it should be an april fools joke because searching for easter related messages isn't exactly something nobody wants to do

... room 0, negative ID? ..

1:26 AM

@Ｕｎｉｃｏｄｅ-ｏｎｌｙ why should string literal always be reference counted? If I do print('static string') that doesn't need to be

Ｕｎｉｃｏｄｅ-ｏｎｌｙ

@Downsheep hmmmmmmmmm ok are they read straight from disk

? nothing would be read from disk in that exampel

Those kind of string literals should go in .rodata

1:55 AM

o/

just almost choked to death on an egg roll after seeing an IBM ad bragging about how they're using The Blockchain to track tomatoes ("from farm to table""

3

Esolanging Fruit

2:28 AM

I just spent more time on Philosophy.SE... it's horrible

The lowest-voted question is "Are we living or dreaming?" and the 5th highest-voted question is "How does one know one is not dreaming?"

The entire site needs to be closed as "too broad." The kinds of questions allowed there seem to violate points 3, 4, 5, and 6 of the "Guidelines for Great Subjective Questions" on "Good Subjective, Bad Subjective.".

user image

The entire PPCG site needs to be closed as "not a question". :/

Some sites does not really follow the general SE rules.

Also, that is on stackoverflow.blog.

Esolanging Fruit

@user202729 It makes mostly the same points as the Help Center does for most sites.

If "suitability for the SE network" is a cliff, then most sites are safely on top.
Philosophy is grasping on the edge by its fingertips.

PPCG is so far off the edge that it's grown wings, so it's fine.

2

The parentheses on the starboard is unbalanced. Twice. ))

2:47 AM

@quartata could be worse, you could have choked on a cherry tomato

79

Q: We're not a Q&A site. But what should be done about it?

It has come up several times recently (more than usual) that PPCG differs from most of the other Stack Exchange sites in that it's not a Q&A site. People don't come here to ask a question because they have a problem, people come here to solve recreational challenges. The most valuable contributio...

discussion site-health

Destructible Lemon

3:01 AM

@user202729 if i were to star this it would become unbalanced 4 times

@quartata Is it an April Fools?

okay problem, arguments are not passed as pointers

so that means it doesn't make sense to be able to reassign arguments

so @Ｕｎｉｃｏｄｅ-ｏｎｌｙ should we 1) put the function arg in ptr 2) not allow

that doesn't mean a function will take i32* instead of i32. we'll just allocate space for the arg on stack

@Downsheep no, it was a real ad

google it

Esolanging Fruit

CMP: If I have an (admittedly rather long) submission to a complex challenge and I post it as multiple answers, gaining me more upvotes and (possibly) bounties, is that abuse?

i think the only time that's ok is if it's something that legitimately has multiple parts, like QFT. Not because of rep, just for presentation purposes

Ｕｎｉｃｏｄｅ-ｏｎｌｙ

@Downsheep 1 (not for primitives though of course)

Esolanging Fruit

3:15 AM

@quartata I never said it was for more rep, but that would certainly be the outcome (if QFT is any indicator)

Also, this wasn't hypothetical. I was trying to get people to agree to flag QFT as "not an answer."

You have foiled my plan

Ｕｎｉｃｏｄｅ-ｏｎｌｙ

@EsolangingFruit yes if you could fit it in one post

part of the reason why qft is more than one answer is because it wouldn't fit in one post. another part is because it's such a massive answer multiple people have had to put in significant work for it

@Ｕｎｉｃｏｄｅ-ｏｎｌｙ The problem is caused with primitives though

Ｕｎｉｃｏｄｅ-ｏｎｌｙ

@Downsheep you want a = 9 to change the var passed in?

Esolanging Fruit

@Ｕｎｉｃｏｄｅ-ｏｎｌｙ I was under the impression the solution for that was supposed to be an external link providing more detail rather than breaking into multiple answers.

But I can't find any relevant meta/help center info

Ｕｎｉｃｏｄｅ-ｏｎｌｙ

@EsolangingFruit no...

@EsolangingFruit external links might break

@EsolangingFruit plus all research was done for the question. external links are basically just for a more comprehensive (pre-existing) guide that the answerer probably referred to

3:19 AM

@Ｕｎｉｃｏｄｅ-ｏｎｌｙ phrased better: should arg refs be immutable?

Esolanging Fruit

Really the QFT answers are still an overview, though—it doesn't contain the actual solution.

@EsolangingFruit good luck with that, it'll take 10 delete votes to delete the question or 18 to get rid of each answer individually

doing that would change a in terms of what it refers to

Ｕｎｉｃｏｄｅ-ｏｎｌｙ

@EsolangingFruit because the actual solution wouldn't fit in any number of posts

@EsolangingFruit It doesn't? Really? ...

Ｕｎｉｃｏｄｅ-ｏｎｌｙ

3:20 AM

@Downsheep idk, but doing arg = foo should not change what the arg passed in refers to

Question to all TNB: Should argument references be immutable

Esolanging Fruit

@Downsheep No. It's annoying.

not if they have inout

Esolanging Fruit

There's no reason they shouldn't.

or whatever your equivalent is e.g. Prolog's modes

Esolanging Fruit

3:21 AM

@user202729 It has the score, but not the actual Life starting state (nor does it include a direct link to some file containing it).

All the answers on QFT includes images.

The images can be used to reconstruct the computer.

Ｕｎｉｃｏｄｅ-ｏｎｌｙ

@user202729 images != solution...

@user202729 it's not easy...

@Ｕｎｉｃｏｄｅ-ｏｎｌｙ "image of solution" == "solution" ...

Esolanging Fruit

@Ｕｎｉｃｏｄｅ-ｏｎｌｙ So there's going to be an external link either way; it's just a matter of how much detail you put in. And they put in too much detail, requiring them to split up the answer.

@quartata We'd take it to Meta, and if the proposal was well-received a mod could hammer it

In any case, I'm not serious about it anyway.

You literally just have to read the pixels, convert it to a suitable format, and then metafy it.

Ｕｎｉｃｏｄｅ-ｏｎｌｙ

3:23 AM

@EsolangingFruit otherwise people won't be able to understand it?

@user202729 yes. but this is ppcg (programming puzzles and code golf), the only time we allow images as solutions is if they're required (e.g. piet/braincopter)

@quartata ok cool

Ｕｎｉｃｏｄｅ-ｏｎｌｙ

@user202729 yeah none of which are easy. if it's that easy then why don't you write one now >_>

The challenge does not require any specific program format. Images are acceptable.

Ｕｎｉｃｏｄｅ-ｏｎｌｙ

@user202729 that's true, but it isn't preferred

> 762 557 533 512 501 482

Esolanging Fruit

3:27 AM

Bottom line is that each of the answers taken by itself aren't solutions. The answers taken collectively aren't a solution, or at least are a low-quality solution solution since it doesn't make the actual submission (initial state) clear (and it seems like it would take a bit of work to actually get to it).

Yes, I just called QFT low-quality. Deal with it.

406 minutes makes a large difference...

Ｕｎｉｃｏｄｅ-ｏｎｌｙ

@user202729 O_o

good guy firefox: has a special mobile browser that clears your stuffs whenever you close it

Ｕｎｉｃｏｄｅ-ｏｎｌｙ

@Picard wait wat

@Picard also is this not what incognito is for

@Ｕｎｉｃｏｄｅ-ｏｎｌｙ yes, but this is literally an entire app for that

firefox focus

also blocks trackers, dunno if incognito does that by itself (without extensions)

Ｕｎｉｃｏｄｅ-ｏｎｌｙ

3:41 AM

doesn't brave browser basically do that too

3:54 AM

no clue

New Sandboxed Posts

0

A: Sandbox for Proposed Challenges

Sandbox Notes Which tags should go? Is this uncomputable? Should I add more test cases? Solve logical statements about arithmetic math logic number number-theory arithmetic integer decision-problem code-golf Ok, some of these need to go. Your task is to determine the truth ...

0

A: Sandbox for Proposed Challenges

Inversion languages cops-and-robbers language-design Cops For this challenge you will design two languages, A and B. Both A and B should be Turing complete. When input into your two languages a program has four options: It halts in both A and B. It halts in A but not B. It halts in B but n...

Anonymous

@EsolangingFruit Yes, it does - right before the line break.

Not self contained. But AFAIK external links to github are allowed (?)

Anonymous

The last word in my answer before the linebreak is a link to this large file, which is a gzipped macrocell file containing the starting pattern (in Life rules, not VarLife) for the computer with Tetris loaded.

Anonymous

@user202729 We prefer to have answers be self-contained when possible, but in cases like QFT where it's not possible (due to answer length limits), a link on a site unlikely to go down anytime soon is the best that can be done

4:02 AM

> Is this uncomputable?

@EsolangingFruit Yes.

Anonymous

@EsolangingFruit It is very clear, and it takes very little work other than reading and loading a file into Golly.

@EsolangingFruit ...

Anonymous

Though there is some valid criticism there - we should probably put the link to the pattern file in the overview

@EsolangingFruit Suggested test case (good luck getting an answer)

Esolanging Fruit

4:32 AM

@user202729 I wonder whether a brute forcer for this is TC then...

The output can only be "yes" or "loop forever".

How can it even "output"?

the only output requirements for being TC are to be able to halt and not halt, no additional output is required

If the problem is uncomputable I do believe that a brute forcer would be TC.

... that qualifies as TC?

I don't even see any meaningful way to get the output.

Output is just a frill added to modern languages to make them easier to debug.

Esolanging Fruit

4:51 AM

Halting? Ha ha ha, I use Three Star Programmer. My entire program is an infinite loop!

Do you think my regex complement challenge is ready to post to Main? (it's been in the Sandbox for four months)

Yes.

5:08 AM

0

Q: Complement of a Regex

Complement of a Regex code-golf regex For the purposes of this challenge, we define a postfix regex dialect with alphabet {1, 0} and the following operations: 1 and 0 match themselves literally. _ matches the empty string. ! always fails (i.e. it does not match anything). ab; matches a, follo...

code-golf parsing regular-expression

Esolanging Fruit

Posted.

For once, I've created a difficult challenge that I know to be possible.

what should the behavior of cast(signedInt, to: unsignedInt) do? @Ｕｎｉｃｏｄｅ-ｏｎｌｙ

Should it just chop off twos complement bit

@Downsheep Add 128.

Anyone know why this old popularity-contest of mine on Approximating a Bell Curve has been receiving a number of answers today/yesterday?

Might have just been bumped to the front page because of the first answer, but I wanted to know if there was something special going on

5:24 AM

@Justin No idea.

5:47 AM

CMC: create a C function that, given an integer (unsigned 32-bit), outputs it to STDOUT, no malloc() or other dynamic memory allocations allowed

stop getting chat to implement your runtimes

@Ｕｎｉｃｏｄｅ-ｏｎｌｙ btw what is integer division operator in VSL

...you know, making a challenge to make a small compliant libc would be pretty funny

@quartata VSL doesn't have a runtime beyond crt0 and the vsl.upref/downref functions

uh

why aren't you naming them upgoat/downgoat

rookie mistake

5:53 AM

@quartata crap brb

every GC I know names them that, it's just the terminology

nvm apparently they are inlined so they don't have names

rip

you can force them to not be inlined. I'd say it's a worthy sacrifice

although.... If you build it as a dynamic link library the symbols will still be there

I take it this is statically linked in

@quartata yeah, right now there isn't support for importing another VSL module as a dylib

missing forward declares?

I mean, you're not alone

5:57 AM

VSL doesn't have forward decls. But right now if you do funcFromAnotherModule, the backend will go ahead and immediately generate funcFromAnotherModule rather than create an external decl

Go doesn't do any dynamic linking

Whether that's a good or bad thing is up for debate

@quartata you can't even manually specify each function that is in dynamic?

@Downsheep ... why do you even need malloc?

Oh, I was wrong. New in 1.5

@user202729 You don't need it but I'd imagine some people want to make sure they allocate exactly the right amount of chars

6:02 AM

I mean, if you use glibc I/O it's just going to go into a buffer anyways

no harm in going a digit at a time and flushing

Does anyone know if _Znwm (C++s heap alloc init) generally maps to malloc() or if it does its own thing

@Downsheep ... what is that? Which compiler? ...

@quartata clang

operator new gets overridden so much in production code I would be hesitant to assume anything

(why?)

just curious

6:09 AM

@user202729 it's the mangled form

@Downgoat you can pick through the source if you're curious about implementation (I don't know myself):

github.com/gcc-mirror/gcc/tree/master/libstdc%2B%2B-v3

have fun! don't get eaten by any code grues!

actually GCC source code is fairly decent

most stl implementations i've looked at use strrev for itoa which I don't want to do though since I don't have generic initializers in VSL yet (and therefore I can't do Pointer<UInt8>(alloc:) so therefore I can't make dynamic CStrings)

oh wait you are talking about _Znwm

looking like it does use malloc

not surprised

libstdc++ is supposed to be a layer on top of your libc, for the most part

@Downsheep that's just to reverse the digits at the end. if you calculate the log of the number first you'll know how big the string is and can build it in reverse

wouldn't recommend but it's probably not much worse

@quartata I still have to build a string though which I can't do :P

I don't know how these libraries are allocating stuff

@Downsheep oh. I thought you meant, like, reallocating

Well...

If your concern is dependence on libc (is it?)

@quartata I'm okay with using libc things as long as they are standard. itoa isn't unfortunately and I think glibc implements it as a macro so yeah...

6:20 AM

Oh, I thought you didn't want to link it

anyways I've always wanted to see a language runtime use this:

I am definitely linking libc because no way I am reimplementing malloc :P

Slab allocation

Slab allocation is a memory management mechanism intended for the efficient memory allocation of kernel objects. It eliminates fragmentation caused by allocations and deallocations. The technique is used to retain allocated memory that contains a data object of a certain type for reuse upon subsequent allocations of objects of the same type. It is analogous to an object pool, but only applies to memory, not other resources. Slab allocation was first introduced in the Solaris 2.4 kernel by Jeff Bonwick. It is now widely used by many Unix and Unix-like operating systems including FreeBSD and Linux...

ignore the stuff about kernel objects

the idea is just that if you have a block aligned to a type already, reuse it

I was thinking of doing something like that as part of an AMM opt pass in scopes.

@Downsheep oh it's not so bad. Reimplementing the math functions is much worse

And threads

as much as I roast pthreads they're still a lifesaver

windows doesn't have pthreads though so can't use them :(

6:25 AM

Cor, it looks like malloc already maintains slabs for small types

How do they do that magic

Damn

6:46 AM

Anyone have suggestions for int division operator

@Downsheep ÷?

preferably one that's ASCII

I think that's mapped to / in tokenization stage

@Downsheep %?

that's mod/remainder

@Downsheep Do you have a list of occupied and available symbols?

6:58 AM

@Downsheep If / is float div, \ could be int div. ; is also pretty OK imo

@Adám github.com/vsl-lang/VSL/blob/upgrade-generics/src/vsl/parser/…

@Emigna huh I've never seen this but it seems pretty logical

@Downsheep I can't tell which are (un)used. What is passThrough?

@Adám everything that is in a string is used

(passThrough just tells the tokenizer to add it to the token list without modification)

@Downsheep Since you use bi- (and tri-) glyphs, how about //?

problem is that will be parsed as comment :P

7:04 AM

@Downsheep %%?

@Downsheep Isn't \ used to escape stuff?

@Adám only within strings

8:00 AM

Does J have special notation for powers of 10? For example, can 2*10^6 be shortened?

-1

Q: SQL Select spanning N:M relationship

I have this N:M relationship: CREATE TABLE auth_user ( id integer NOT NULL PRIMARY KEY, username character varying(150) NOT NULL UNIQUE ); CREATE TABLE auth_group ( id integer NOT NULL PRIMARY KEY, name character varying(80) NOT NULL UNIQUE ); CREATE TABLE auth_user_groups ( ...

code-golf tips sql

@Downsheep It's not April Fools anymore, please stop scaring me.

@user202729 Yes, 2e6.

8:36 AM

-1

Q: Is m[i / l][i % l] the same as m[0][i]?

It's probably not but i would like to understand why. If m[0][i] is the same as *(*(m + 0) + i) and the elemnets of m are contiguous in memory, I thought it might be equivalent to m[i / l][i % l]. (l - line length)

code-golf c c++

8:54 AM

If I see a low-quality post that happens to be invalid should I delete?

(this)

I think delete, but the fact that it's short is irrelevant to whether it's invalid...

And getting review badge this way is ...

Why is stdin copied to stdout here?

Looks like TIO is evaluating them...

Because console.

> `type MutableRequired<T> = { -readonly [P in keyof T]-?: T[P] }; // Remove readonly and ?`
> `type ReadonlyPartial<T> = { +readonly [P in keyof T]+?: T[P] }; // Add readonly and ?`

TypeScript's type system is getting a bit silly.

9:18 AM

J: Is there a simpler way to do this?

(assume each row has exactly 2 numbers)

CMC: Given two function f, g, of type (nat -> bool) -> nat, decide whether they are equal

What is "nat"?

Natural integers?

9:41 AM

J: Is there any way to get a tacit verb from its string representation?

".?

128!:2?

Is there a dyadic equivalent of 128!:2?

caird coinheringaahing

9:58 AM

19 hours ago, by caird coinheringaahing

yesterday, by caird coinheringaahing

Suggestions for memory models for an esolang? Specifically for this

Still haven't got any suggestions

« first day (2621 days earlier) ← previous day next day → last day (2212 days later) »

Transcript for

Apr '183

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