The Nineteenth Byte

Business Cat

17:00

If you do int* loc=b; it works fine

Leaky Nun

@Cowsquack that's above my paygrade

@BusinessCat why doesn't it work if I do int loc = b; int *c = loc;?

Cows quack

I'm trying to print what's in the second byte of the int but am failing

if it's even possible

Phoenix

Solution: Make everything a void*

How does pointer arithmetic work on void*? What does it consider the size of a void?

Business Cat

Solution: Don't use C

Cows quack

what is a void *? Is it just typeless?

DJMcMayhem

17:03

@LeakyNun because loc is an int, not an address. You want &loc

@Cowsquack yes

Business Cat

@Phoenix 1 byte

Leaky Nun

@DJMcMayhem what if I want to treat it as an address?

Phoenix

@Cowsquack It's like Object in C#/Java. You can store whatever you want in there.

Cows quack

okay

DJMcMayhem

@LeakyNun (I haven't tested this) You might be able to do int *c = (int*)loc, but that's a terrible terrible idea.

If that doesn't work, you should look into memset

Leaky Nun

17:04

it still segfaults

DJMcMayhem

Or unions which is easier

@LeakyNun Yeah, I'd expect that

Leaky Nun

@DJMcMayhem no I don't want to set the memory

DJMcMayhem

Wait, so what exactly do you want to do? You have an int, and you want to use that value of the int as the address of a new pointer?

Leaky Nun

@DJMcMayhem yes

DJMcMayhem

Is this for something legitimate? Because I don't know what you're doing but I'm positive there's a better way

Business Cat

17:05

That's a segfault waiting to happen

2

Phoenix

@Phoenix although technically, C# does have void*s...

DJMcMayhem

Either way, that's not really going to work since an int* is one byte and an int is (usually) 4 bytes.

Leaky Nun

@DJMcMayhem no way

when I print an int* it gives me 864298784

Phoenix

The size of an int isn't specified, short is guaranteed to be 2 bytes and long is guaranteed to be 4 bytes. iirc.

Cows quack

that's a long byte

DJMcMayhem

17:07

Wait, ignore me

Sorry, I don't know why I said an int* is one byte

That's obviously false

Phoenix

On a 32 bit system, an int* is 4 bytes, on a 64 bit system, it's 8 bytes.

Cows quack

so int* is a pointer that points to an int?

Phoenix

Yes.

Leaky Nun

@Cowsquack yes

Phoenix

Also known as an int array.

Leaky Nun

17:08

(insert mandatory pointer-is-not-array reference)

DJMcMayhem

int b = loc;
int *c;
memcpy(c, &b, 4);

Cows quack

so it points to a location

DJMcMayhem

Let the records show that I'm against this horrible code

Phoenix

Yep

DJMcMayhem

Although, 4 is not really correct, cause it depends on architecture

Leaky Nun

17:09

@DJMcMayhem so sizeof(int*)?

DJMcMayhem

Yes

@LeakyNun Alternate approach:

Cows quack

that means TIO is on a 64 bit system

DJMcMayhem

    union Abomination {
        int loc;
        int *pLoc;
    }

    Abomination dont_do_this;
    dont_do_this.loc = loc;
    int *c = dont_do_this.pLoc;

Phoenix

@Cowsquack Yes

DJMcMayhem

How do you like my var names? :P

Phoenix

17:11

You can tell because Assembly (as, x64, Linux)

Leaky Nun

@DJMcMayhem still segfaults

DJMcMayhem

6 mins ago, by Business Cat

That's a segfault waiting to happen

That's because you probably aren't allowed to touch wherever the value of loc happens to randomly point at

Also, what is int c = b;?

Riley

@LeakyNun d (usually won't) point to a valid address.

Leaky Nun

@Riley but logically it would

DJMcMayhem

Why?

Riley

17:14

@LeakyNun It isn't initialized. In this case, it's probably 0.

Leaky Nun

@DJMcMayhem just look into my code

@Riley I actually initialized it

Riley

@LeakyNun Where?

Leaky Nun

@Riley b[0]=2?

Riley

@LeakyNun That has nothing to do with d.

Leaky Nun

@Riley but d is c which is b

DJMcMayhem

17:16

@LeakyNun I have. I don't quite get it. What is int c = b supposed to do?

c != b because b is a pointer, and c is an int

Leaky Nun

I cast it

DJMcMayhem

No you didn't

Riley

You are copying into the address that d points to, but you never set the address that d points to.

DJMcMayhem

@LeakyNun Ooooooh, you want memcpy(&d,&c,sizeof(int*));

That's right

Dennis

@Cowsquack Shocking.

DJMcMayhem

17:18

memcpy(d,&c,sizeof(int*)); is copying into the location of d (unassigned), not the pointer itself

@Riley (thanks)

Riley

@LeakyNun If you allocate memory for d it doesn't seg foult

Leaky Nun

@Riley no, I want d to be equal to b

DJMcMayhem

Try it online! seems to do what you want (notice that the last line still segfaults though)

Riley

@LeakyNun Oh, then do what DJMcMayhem said.

Leaky Nun

@Riley then it segfaults

DJMcMayhem

17:21

Of course it does

Dennis Ritchie would be turning over in his grave.

Riley

I guess I'm still not quite sure what you're trying to do.

caird coinheringaahing

Can someone have a look at this? I'm thinking of posting and want a last minute check

Poke

Elon Musk's new website

Phoenix

._.

totallyhuman

ಠ_ಠ

Business Cat

17:35

@Poke something.com

Poke

@BusinessCat at least that one has html

lulz

Leaky Nun

CMC: 1^2 + 2^2 + ... + n^2

DJMcMayhem

Given N?

Leaky Nun

sure

DJMcMayhem

:U in MATL

And R² in jelly I think?

Leaky Nun

17:42

@DJMcMayhem what sorcery is this

@DJMcMayhem no it wouldn't work

DJMcMayhem

Oh sorry, I misunderstood

@LeakyNun :Us

Business Cat

R²S is Jelly

totallyhuman

english: range, vector square, sum

Riley

@LeakyNun 05AB1E, 3 bytes: LnO

Leaky Nun

try to do it... in languages where this wouldn't be trivial

DJMcMayhem

17:43

That would be a great challenge for brain-flak

Leaky Nun

@DJMcMayhem no it wouldn't

DJMcMayhem

I'll do that in a few minutes

@LeakyNun Totally would

Leaky Nun

it's just a cubic polynomial

totallyhuman

well if i manage to implement range and map in ,,,: R[2*]#

Leaky Nun

do it in oasis

totallyhuman

17:45

lambda n:sum(i**2for i in range(1,n))

Leaky Nun

@totallyhuman you can get smaller

Business Cat

@totallyhuman In Gaia I made range and map but I haven't made a n^2 operator yet so I guess I have the opposite problem ¯\_(ツ)_/¯

totallyhuman

if it involves crazy math i have no idea :P

Leaky Nun

@BusinessCat just multiply n by itself

2 mins ago, by Leaky Nun

it's just a cubic polynomial

Business Cat

I also don't have a sum operator yet so I'd have to reduce by addition

Leaky Nun

17:49

@BusinessCat what is Σ?

oh, notImplemented

Business Cat

It going to be sum, I just haven't made it

I could do it in 5 bytes for now: :×†+⊢

Riley

@LeakyNun Brain-Flak, 36 bytes: ({<(({}[()])())>{({})({}[()])}{}}{})

Leaky Nun

@Riley nice

Phoenix

Sum could be unary addition on lists.

DJMcMayhem

34: Try it online!

Business Cat

17:50

36 is a lot shorter than I expected it to be

DJMcMayhem

I bet it could be shorter still

Leaky Nun

4 mins ago, by Leaky Nun

2 mins ago, by Leaky Nun

it's just a cubic polynomial

Business Cat

@BusinessCat When I actually make these operators it'd just be s¦Σ

Leaky Nun

CMC (language-specific): write some code so that + would be evaluated as *

totallyhuman

#define + * i think

Leaky Nun

17:53

language specific because it might not be possible in some languages

Business Cat

Evaluated as * normally is in that language?

totallyhuman

i don't know c though

Leaky Nun

@BusinessCat I mean, addition as multiplication

Phoenix

At least one language out there, +=*. plsplspls

totallyhuman

@totallyhuman is that right? anybody who can c?

DJMcMayhem

17:54

o/

Phoenix

int operator +(this int a,int b)=>a*b works in C# I think maybe?

DJMcMayhem

It's right

Business Cat

Oh yeah, in Scheme you can do (define + *)

Leaky Nun

@DJMcMayhem but it would produce side effect like +a being treated as a pointer?

DJMcMayhem

Yes

totallyhuman

17:55

lol

DJMcMayhem

It would scan through the whole file before compiling, look for any occurences of +, and replace it with * before getting to the compilation stage.

totallyhuman

in python you'd probably just redefine int.__add__ and other classes

int.__add__ = int.__mul__

Peter Taylor

@Phoenix No, you can't define extension methods for operators

totallyhuman

aw it doesn't allow it

Business Cat

TypeError: can't set attributes of built-in/extension type 'int'

DJMcMayhem

17:57

Python 3.6.0 (v3.6.0:41df79263a11, Dec 23 2016, 08:06:12) [MSC v.1900 64 bit (AM
D64)] on win32
Type "help", "copyright", "credits" or "license" for more information.
>>> int.__add__ = int.__mul__
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
TypeError: can't set attributes of built-in/extension type 'int'
>>>

totallyhuman

you both were ninja'd

DJMcMayhem

I know

Leaky Nun

imaginary cookie for anyone who does it via java reflection

Jim

@LeakyNun Pyth: D+GHR*GH

HyperNeutrino

@LeakyNun I think Java doesn't allow that, but I'm not sure.

Leaky Nun

17:58

@Jim hmm, nice

Phoenix

@LeakyNun I do not believe that to be possible without modifying the source and javacing it.

Leaky Nun

@HyperNeutrino I thought everything is possible in java reflection

HyperNeutrino

Well not redefining variables

Phoenix

You can't do normal operator overloading with Java reflection.

HyperNeutrino

because addition in Python does a + b => a.__add__(b)

but a + b in Java is just a + b

Leaky Nun

17:59

but you can change 4 into 5, lol

Transcript for