/dpt/ - Daily Programming Thread

scamming people out of their buttcoins, reached 1000$ in two days
people are so gullible

>>58368880
That didn't happen.

File: noirtouchscreen.jpg (95KB, 2448x2076px)

95KB, 2448x2076px

Made a camera touchscreen GUI for a raspberry pi camera. It's for an infrared-sensitive NoIR camera.

What other features should I add?

>>58368889
dont worry man, one day you'll make money and be succesful too :)

>>58368880
Post rough summary of methodology.

File: mmm.png (139KB, 379x405px) Image search: [Google]

139KB, 379x405px

what language is this?

>>58368905
memesnek

>>58368905
langoo of typie spacies

>>58368905
Looks like straight up py.

Got tired of writing loose files for every network.

So, making a prompt/interpreter thing for loading/making/saving/training neural nets.

>>58368905
Cobra

>>58368903
if i tell everyone will do it and ruin it, sorry :/

>>58368947
fake news? falsified but believable buttcoin market facsimiles? just plain old social engineering? give us a hint.

>>58368963
alright, here's the hint, imagine yourself as something legitimate and act like it

>>58368977
people really are fucking gullible

>>58368987
indeed, they'll believe everything as long as you make it look as legit as possible

Writing a metaheuristic for solving a network function virtualization problem for my college.

I really hate people who say lisp is implementation of untyped lambda calculus and Haskell implementation of typed lambda calculus.

That's like saying all imperative languages are implementations of Turing machine.

While it may be technically correct it's a huge understatement.

>>58369020
you're wrong.

>>58369036
Why?

>>58368947
If you can't even post a summary enough to give an idea, you're full of shit.

>>58369061
It's an axiom on the level of Haskell actually being useful.

File: 1483453233544.png (818KB, 1280x719px) Image search: [Google]

818KB, 1280x719px

Functional programming.
Last /fpt/: >>58332039

Resources:
>Erlang
http://learnyousomeerlang.com/content
>>Elixir
http://chimera.labs.oreilly.com/books/1234000001642/index.html
http://elixir-lang.org/learning.html
>F#
http://fsharp.org/learn
>Haskell
https://en.wikibooks.org/wiki/Haskell
https://0x0.st/pbp.pdf
>Lisps
>>Common Lisp
http://gigamonkeys.com/book
http://www.cs.cmu.edu/~dst/LispBook/book.pdf
Paul Graham's ANSI Common Lisp
On Lisp
Common Lisp Recipes
Land of Lisp
An Introduction to Functional Programming Through Lambda Calculus
>>Clojure
http://www.braveclojure.com/foreword/
The joy of Clojure
>>Scheme
SICP
Essentials of Programming Languages
How to Design Programs:
http://www.ccs.neu.edu/home/matthias/HtDP2e/
Art of the Propagator
Little Schemer
The Seasoned Schemer
The Scheme Programming Language by Kent Dybvig
Realm of Racket
Lisp in Small Pieces
>OCaml
https://realworldocaml.org/
http://ocaml.org/learn/tutorials/
>Scala
Functional Programming in Scala (Chiusano and Bjarnason)
Atomic Scala (Eckel and Marsh)
Programming Scala (Wampler and Payne)
Programming in Scala (Odersky, Spoon and Venners)
>Web languages
>>Elm
https://guide.elm-lang.org/
>>PureScript
http://www.purescript.org/learn/

is functional programming a meme?

File: 1459231731880.jpg (20KB, 317x265px) Image search: [Google]

20KB, 317x265px

>>58369151
>>58369150

>>58369150
I'm glad that /fpt/ is dead, as less needless generals are better, but you don't need to spam those stupid links here.

>>58369150
>useless programming

>>58369166
Those links will be posted with every /dpt/ (since /fpt/ is proclaimed dead) so every willing to learn functional programming will be able to check this list.

>>58369150
>SICP
Should I read it if I'm interested in FP but not necessarily in Scheme?

>>58369166
Arguably /wdg/ keeps the designer hipsters and pajeets contained.
>>58369151
It has its purposes but is useless for most tasks.

>>58369199
/dpt/ has never had any OP-pasta, aside from "what are you working on, /g/".
Lets keep it that way.

If someone wants to learn programming, functional or not: they can fucking google that shit.

>no side-effects in pure functional languages
>you have to incur the overhead of a recursive function call every single time you want to iterate through something
am i missing something here?

>>58369224
>>58369199

I don't mind it posted here, but maybe add it to installgentoo wiki

>>58369233
>incur the overhead of a recursive function
>am i missing something here?
Almost all functional languages have tail-call optimisation.

>>58369201
You might want also check this: >>58345943
SICP alone is not much helpful in learning FP.

>>58369233
Most functional languages optimize tail-call recursion into loop (with go-tos) when compiled.

Also most languages aren't pure, but even in pure languages you can use side-effs with monads.

And no, monads aren't some hack to hide away side-effs, they are pure.

>>58369245
>>58369260
guess i was missing something pretty major. should have just let me talk myself out of learning functional programming

>>58369224
No one posts it in /dpt/ OP.

Can someone redpill me on functional programming? How do lambda functions improve programming skills?

>>58369250
I'm not sure I can get them right now. I prefer to have books like these in print.
>SICP alone is not much helpful in learning FP.
Is it still good though? I'm not that experienced with imperative programming either.

>>58369288
It is fine.

>>58369302
I don't need to know much about Scheme, right? I already know basic functional concepts so that shouldn't be an issue

We need /prog/ to come back. Also, make it an imageboard so people actually have a reason to give a fuck about it.

>>58369285
Lambdas aren't exclusive to FP.

>>58369285
Lambda functions by itself aren't that powerful.

Functional programming isn't great because it has x feature, but because it has many concepts which interact together and is more than just sum of it's parts.

For example lambdas by itself are sometimes convenient, but when combined by higher order functions they allow you to specialize functions (like sort, filter, map) and to generate new functions at run-time for your needs. When you add closures to the mix you get the ability to define new data structures by functions and abstract away mutability.

In short, functional programming is great and will improve your skills by giving you an entire new toolbox, which you will be able to use in other languages too. For example, in scala, ruby, lisps, you will be able to specialize object methods at run-time, generate new sorts of objects, etc.

If nothing else you will finally understand recursion, which is needed pretty much everywhere, and why it's smart to abstract it away and use maps, reduces, folds, ... instead.

>>58369313
SICP assumes no prerequisite knowledge.

if i'm understanding tail-call recursion optimization at all, there's negligible extra instructions at the assembler level. what is it that prevents C and C++ from optimizing for tail recursion? what's the opportunity cost?

i don't understand exactly how it's implemented, but the first thing that seems like it'd be an issue is evaluating whether or not a terminating jump statement references its own function's memory address. seems like there could be significant overhead in languages not expressly designed for tail recursion.

help me out here.

I want to do threading and mutex locks in a C++ program on Windows.
The simplest option is to just use Visual Studio, yes?

>>58369313
You do not need, SICP implies students don't know Scheme.

>>58369337
>what is it that prevents C and C++ from optimizing for tail recursion?
Side effects

>>58369352
so i'm not understanding it at all. back to reading.

>>58369337
Most of compilers for C/C++ optimize it though.

>>58369352
Irrelevant, this optimization basically replaces function call with go to. Most of lisps (which indeed all implement it) have a lot of side effs.

>>58369337
>what is it that prevents C and C++ from optimizing for tail recursion?
Many compiler C and C++ compilers do TCO, if the situation is correct.

>>58369359
Don't worry he has no clue what he is talking about.

>>58369366
correct
http://stackoverflow.com/questions/34125/which-if-any-c-compilers-do-tail-recursion-optimization

>>58369337
> what is it that prevents C and C++ from optimizing for tail recursion?
Nothing? https://godbolt.org/g/XNuQN6

File: asuka.jpg (474KB, 1619x1725px) Image search: [Google]

474KB, 1619x1725px

Scala vs Haskell ?

>>58369380
neither

>>58369380
giant turd vs diaerrhea

>>58369380
Haskell ofc, subtyping is an unnecessary complication of a type system.

>>58369387
do you want a bloody pucker hole or a messy cleanup?

>>58369390
>MUH TIGHP FEEREE

>>58369285

Functional programming is not just about lambdas. Those are just a useful tool generally provided by functional languages, and sometimes provided by non-functional languages.

Idea behind functional programming is to use the function as the basic unit of abstraction, and that functions should behave more like the mathematical definition of a function, rather than procedures, which are allowed to have side effects. A function receives values as input and returns a value, and it can be guaranteed that every time a function is called, the result is the same.

Of course, you generally don't go absolutely without side effects. A program with no side effects can be considered equivalent to either an empty program or an infinite loop, since I/O is a type of side effect. Rather, when side effects are used, they are explicit. General idea is to make debugging a hell of a lot easier when you can't have each and every function changing whatever it fucking wants.

A functional language is a language designed to be used with the FP paradigm. To do so, they typically make functions into first class objects, allowing them to be passed around as values and returned from other functions. They also tend to make things like immutability the default, and require the programmer to be explicit whenever they want to so much as change a variable (i.e. let mut vs let in F# and Rust).

how can you write a tic tac toe game elegantly without checking a billion conditions manually?

>>58369396
how does functional programming deal with the need to get closer to the chrome? is the ideal to abstract away everything?

>>58369407
you cant

its just x and o, what elegance do you need, it\s literally two functions

>>58369418
All languages provide means of combination, means of abstraction and primitives.

Functional languages have the best means of abstraction and combination, while most of their primitives are higher level, pretty far from metal.

If you need a language close to metal with good means of abstraction try forth. It's not as good as functional languages, but way better than imperatives & oops.

>>58369407
your only relevant conditions are x's or o's which are touching each other, specifically those which are in a straight, unabridged line. you might need to eventually delve into checking the shit out of every possible thing but you can probably do away with brute forcing the initial stages of the game by just checking if anything is connected to something of its own type

>>58369407
sum neighbors of 2, 4, 5, 6, 8, orientation at a time, if any of them =2, end game?

>>58369337

>what is it that prevents C and C++ from optimizing for tail recursion? what's the opportunity cost?

Nothing. No part of the C or C++ standard requires explicit stack frames to be created on each function call, and most popular compilers perform aggressive optimizations, including but not limited to tail call optimization. That said, no part of the standard requires them to be performed either. I believe the C standard writers wanted the language to be easy to implement, and the C++ standard writers just don't give a fuck about anything.

>>58369418

Programming paradigms are not one size fits all solutions. They are an idea for how to design software, but not every program type fits well with every paradigm. If a program by its very nature must be close to the metal, then there's no way you're going to end up sticking tightly to FP principles. At best, you can take everything that DOESN'T need to interact with the metal, and abstract it away to a function.

>>58369407
for each X or O placed, check its neighbors until either three in a row or it hits its opposite/nothing. worst case scenario is six checks, best case is three. most likely performs worse than >>58369461 in a best-case scenario, but realistically most likely does much better. not sure what either algorithm is in big O.

>>58369540
>the C++ standard writers just don't give a fuck about anything.
Indeed. They're based as fuck assholes from the previous century.

>>58369559
unlike you from the current one

>>58368905

looks like python

How do I learn COBOL?

>>58369585
take a time machine

>>58369585
be old

>>58369563
And how would you know? You're right though, Roast Master, I need to be more based!

I'm off to writing a new clusterfuck of a language rightfully named Clusterfuck. It will literally rape Stroustrup for fun.

>>58369600
i know because i had consenseual sex with your female parent

you do that

>>58369600

We literally already have a programming language called Brainfuck.

>>58369585
I'm mildly surprised it has a new revision as of 2014

>>58369654
Well, in 2008 80% of big business still had cobol code. God knows how much of it obeys the standards though.

Cobol isn't going anywhere anytime soon and due to nobody knowing it, knowing it it's the greatest job security.

>>58368901
Maybe one day you'll be normal enough to not lie online.

Here's the 'elegant' tic-tac-toe

f = {
    {false, true, false},
    {true, true, true},
    {false, true, false}
}

u = {
    {0, 0, 0},
    {0, 0, 0},
    {0, 0, 0}
}

o = {
    {{1, 1}, {-1, -1}},
    {{1, -1}, {-1, 1}},
    {{-1, 0}, {1, 0}},
    {{0, -1}, {0, 1}}
}

results = {}

for x = 1, 3 do
    for y = 1, 3 do
        if f[x][y] then
            for _o = 1, 4 do
                if u[x-o[_o][1][1]][y-o[_o][1][2]] and u[x-o[_o][2][1]][y-o[_o][2][2]] then
                    table.insert(results, u[x-o[_o][1][1]][y-o[_o][1][2]] + u[x][y] + u[x-o[_o][2][1]][y-o[_o][2][2]])
                end
            end
        end
    end
end

for i, e in ipairs(results) do
    if e = 3 or e = -3 then
        print("Game Over!")
    end
end

>>58369111
nice try, i still wont give my idea tho

>>58369692
maybe one day your life will be good enough to tolerate others having good ones :^)

>>58369699
>end
>end
>end
>end
>end

>>58369713
>I scam ppl make lot of money
>cant show any proof though sorry xd
Literally the "My dad works at X" meme.

>>58369736
fbi please go

>>58369718
>{
>{
>{
>{
>{
>}
>}
>}
>}
>}

>>58369745
>)))))))))))))))))))))))))))))
still better than end end end end end end end end end end end end end end end end end end end end end end end

>>58369752
I like how you're more concerned about that than this: u[x-o[_o][1][1]][y-o[_o][1][2]]

File: 36c.jpg (84KB, 907x661px) Image search: [Google]

84KB, 907x661px

>>58369742

File: 1471652512201.png (8KB, 687x450px) Image search: [Google]

8KB, 687x450px

>>58369768
lmaoooooo

>>58369752
>>58369745

Emacs paredit-mode is godlike for this.

>>58368865
just making a home page for my browser

File: Screenshot_2017-01-07_12-21-39.png (18KB, 749x190px) Image search: [Google]

18KB, 749x190px

I had to google "sum integers in JS", didn't I. Oh god.

what would be the worst order in which you could learn several programming languages?

>>58369837
In JS that's just:

[1,2,3,4].reduce((a, b) => a + b);

>>58369898
I know. But somebody had to suggest using jQuery for summing integers and 60 people patted him on the back for it. It was a bit of a shock.

>>58369884
COBOL, Fortran, BASIC, Haskell.

// linked list - a collection of nodes "linked" together by pointers. 
// Each node has data and pointer to next or NULL if last.
#include <stdio.h>
#include <stdlib.h>


int bs(void) {
    struct node {
        int data;
        struct node *next;
    };
    struct node *head = NULL; // declaring nodes having pointers point to NULL
    struct node *second = NULL;
    struct node *third = NULL;
    
    head = malloc(sizeof(struct node)); // allocate memory for nodes in heap
    second = malloc(sizeof(struct node));
    third = malloc(sizeof(struct node));
    
    head->data = 1;
    head->next = second;
    
    second->data = 2;
    second->next = third;        // assigning data to nodes and pointing to next node in list
    
    third->data = 3;
    third->next = NULL;
    
    return head;
}

int length(struct node *head) { // essentially lets you iterate through a list with modification
    struct node *current = head; 
    int count = 0;
    
    while (current != NULL) { // do something with current node here in this case inc count to find amt of nodes in list
        count++;
        current= current->next;
    }
    return count;

int main() {
    struct node *myList = buildf();
    int len = length(myList);
    printf("%i", len);
}

}

This retarded piece of shit. I hate being a beginner

File: file.png (32KB, 753x772px) Image search: [Google]

32KB, 753x772px

>>58369913
well yeah that's stackoverflow for you

>>58369643
Make it as popular as Sepples and we will talk.

>>58369837

rubyist@Overmind:~$ nodejs
> var x = [1,2,3,4,5];
undefined
> var sum = 0;
undefined
> for (var i = 0; i < 5; i++) { sum += x[i]; }
15

>>58369925

bs should return struct node* instead of int. length should return size_t instead of int.

>>58369925

Also, struct node should be defined outside of the bs function.

>>58369925
>>58369925

I'm a C noob too and this doesnt look like a desirable way to build a linked list.

am I wrong?

>>58370044
linked lists generally are never desirable

>>58370044
probably this is my first foray into linked lists it's trial by fire

>>58370050
Says somebody who never used lisp.

>>58370068
i've no business using something that's not efficient so you'd be right about it

>>58370009
>2017
>still using for loops
>not for each
pls stop

sum = 0; [1,2,3,4,5].forEach(x => sum += x);

is ok, but >>58369898 is best

>>58370068
Lisp got nice functions for list manipulation but you shouldn't lists just because.
Lisp is general purpose language and has all the most common data structures and you should use the right data structure for your task.

>>58370068

If you want Lisp code that doesn't run like ass, don't use lists that much.

>>58370064
I would do it like some lisps do.

First define dotted pair, that is a datatype with two fields car and cdr.

Now you build a linked list from pairs by putting values into cars and pointers into cdrs.

Now you can build trees by putting arbitrary values and pointers into cars/cdrs.

>>58370076
>>58370064
>>58370086
>>58370094
Common lisp is in some areas quite close to c++ and faster than java most of the time. Lists are useful because they are one of the most diverse data structures and easy to use. Later you can always replace them with other structures, but for prototyping they are invaluable.

>>58370009
thanks for the advice appreciate it

File: xartb.gif (4KB, 390x243px) Image search: [Google]

4KB, 390x243px

>>58370068
Classical Lisp has no liked lists though, it's just lispers was shut in their echo-chamber for 60 years so they keep calling binary trees "lists" for some reason.

>>58370096
> Common lisp is in some areas quite close to c++ and faster than java most of the time.
Can you please stop spreading misinformation based on a single article from 2000, because in 2017 it's just false: https://benchmarksgame.alioth.debian.org/u64q/lisp.html , https://benchmarksgame.alioth.debian.org/u64q/compare.php?lang=sbcl&lang2=gpp .

>>58370079

For a task this simple, it is fine to use a regular for loop, rather than a function.

>>58370096

There are plenty of languages that reach near C++ levels of performance in very specific conditions when writing non-idiomatic code. Linked lists are the idiomatic way to do most things in Lisp. However, they are inherently slow by design.

>>58370107
Linked list can be implemented using binary trees.
And majority of those trees in lisp are lists.

>>58370140
Yes, and? These ares still different things. "Linked list" has a specific meaning in CS for decades now, and refusal by lispers to call thing by their proper names just shows how out of touch with the real world they are.

>>58370138
>For a task this simple, it is fine to use a regular for loop, rather than a function.

Why not just use reduce?

>>58368865
Any ios developers here?

I'm making an app and I never published anything on the app store before. How likely is it that my app will be rejected if I use http rest requests in my app?

Critique please and thank you.

// -1 for X and 1 for O
const board = [
    1, 0, 0,
    0, 1, 0,
    0, 0, 1,
]

function check() {
    const lines = []

    // Horizontal and vertical
    for (let i = 0; i < 3; i++) {
        const hor = [],
            vert = []
        for (let j = 0; j < 3; j++) {
            hor.push(board[3 * i + j])
            vert.push(board[3 * j + i])
        }
        lines.push(hor, vert)
    }

    // Diagonal
    const d1 = [],
        d2 = []
    for (let i = 0, j = 6; i < 9; i += 4, j -= 2) {
        d1.push(board[i])
        d2.push(board[j])
    }
    lines.push(d1, d2)

    for (let l of lines) {
        switch (l[0] + l[1] + l[2]) {
            case 3:
                return "O"
            case -3:
                return "X"
        }
    }
    return ""
}

>>58370140
>majority of ... in lisp are lists
>majority
>in lisp
>in list processor

>>58370138
no, it's disgusting.
it introduces a global variable i, creates the possibility of off by 1 errors and doesn't show its intention as well as reduce or for each

File: theory-programing.jpg (62KB, 874x451px) Image search: [Google]

62KB, 874x451px

>>58369233
Hey hey my friend, you need to abandon that computer slave mindset. You are not C/C++/ASM monkey, you are a developer who solves real-world tasks that bring business/scientific value.
You do not need to bother yourself how exactly computer will run your program - you only declare abstract high-level steps that make sense to the developer, not some "optimized" spaghetti code.

>>58370194

Which would you rather use if you had to call the sum function a million times?

>>58370231

>global variable i
Local to the function that called it, not global.

does solving a ton of programming puzzle actually enable you to make actual contributions to projects?

>>58370305
That's like saying "does solving a ton of sodukus enable me to contribute to number theory paper writing?"

>>58370323

I think that's a poor comparison

>>58370266
Left is with green threading runtime, right is without.

>>58370272
>it's ok for a task this simple
>what if you had to do it a million times
it would be different for big arrays, you'd probably have to make it asynchronous so the loop doesn't block other browser events, but I'd use a for each loop

>local to the function that called
it was global in your case

>>58370272
>Which would you rather use if you had to call the sum function a million times?

I'd refactor because that sounds like a poor design.

File: maze.png (18KB, 462x830px) Image search: [Google]

18KB, 462x830px

Wrote a maze generator for my dungeon crawler then made some code that "upscales" the maze and creates rooms with passages based on it (in case of pic related the maze scale factor is 5).

Probably could reduce the overhead walls on each side but idgaf.

>>58370450
nifty

>>58370406

In a REPL, I have no reason not to spam globals left and right. In a real program, this would obviously be relegated to some function, maybe called sum_elements. Within the scope of a function, it is fair game to use regular old for loops wherever they make sense.

In any case, I suppose a foreach loop would also have been acceptable.

>>58370450
>maze
>1 fork

>>58370450
You need to read this friend: Mazes for programmers

Because your maze is bad and you should feel bad.

>>58370720
What's bad about it? DFS just werks.

>>58370750
One fork in entire maze...

>>58370761
Which is not a problem at all for my purposes, nice try Jamis anyway, but I'm not spending $25 on your book.

>>58370761

So did you notice that it's a 11x11 maze?

>>58370795

Didn't we have a big fucking torrent of programming books lying around somewhere?

>>58370841
The torrent's old as fuck and infected with malware.

>>58370874
>>58370841

On wiki you have the torrent, idk about malware it works fine for me.

For other books (like this maze one): http://gen.lib.rus.ec/

>>58370841
How come this torrent be 35GB?

>>58369020
>While it may be technically correct
But that's the best kind of correct.

Is Rust a good language?

>>58371363
Define good.

>>58371363
It's interesting.
I don't know if it's going to stand the test of time.

>>58371370
Define define.

>>58371370
Can it be useful? Is it fun to use?

>>58371363
it's not, because it doesn't have a lot of jobs.

>>58371398
>because it doesn't have a lot of jobs
That logic is terrible.
Java has a lot of jobs, but it's complete garbage.

>>58371370
>>58371381

#define good "good"
#define define "define"

>>58371407
I disagree. java is alright because it has a lot of jobs. ofc, it could be better, like c++, which is best of both worlds.

>>58371388
Yes.
Yes.

>>58371422
Please, 'argumentum ad populum' is a logical fallacy.
Stop acting like it's not.

By your shitty logic, every new language is complete garbage, just by the virtue of being new.
We will forever be stuck with the shittyness that we have today.

>>58371388
Yes on both, but it is still very immature. The compiler is slow and its optimiser is primitive. The library pool is growing, but needs a few more years at least.

>>58371422
Pajeet plz

>>58371422
That's why he needed to define good. If good means jobs, then Rust is shit and Java is god tier.

>>58368865
Learning c++ memory management since i've been a harlot that relies on built in allocation until now.

>>58371447
It's not argumentum ad populum. One of the variables I use to determine whether it's a good language is if it can help you make a living. java does that very well, so it's a good language because it excels in one of the requirements.
>>58371450
I'm thankfully not one of the fecal folk.

>>58369314
I was thinking this the other day.

File: Screenshot at 2017-01-07 13-33-18.png (416KB, 1366x768px) Image search: [Google]

416KB, 1366x768px

Hi /g/ I have a quick question. I am currently playing with Nokogiri and was wondering how do I scrap all the threads on a board. I am thinking to make a simple minimalistic /g/ thread viewer. I dont really have nothing else than
>page=HTTParty.get('http://boards.4chan.org/b/').
any hints ?

>>58371547
search: vichan api

>>58370176
> "Linked list" has a specific meaning in CS for decades now,
Yes, except that what he said is based on >>58370107 is incorrect, it's actually the other way around and it fits the CS definition perfectly. that Basically a linked list which "contains" other linked lists is a tree. No one said that the data (structure) in a linked list can't be a pointer to another linked list. This essentially results in a tree.

>>58371639
u are so dum read SICP

>>58371802
Ok, thx, bye. I mean you can view it both ways, but still.

>>58371818
for a linked list to contain more than one linked lists it doesn't have to be a tree

a linked list of length 2 contains 3 linked lists

>>58371398
By that logic, Java is the best language

>>58371850
Yes, I know what you mean. You're right.
My point was that you can use a linked list to build a tree. I was mainly referring to the image and similar situations arising in Lisp like ((1 2 3) (4 (56))) , etc. It's a way to build trees.

>>58371859
How could you possibly know that? I specifically said it was only one of a number of factors.

>>58371883
yep that's a tree structure, basically
though there are no parent pointers, cuz singly linked

Visual Studio supports std::thread and std::mutex and C++14, yeah? Tired of minGW lacking every other thing and having to write shitty hacks or deal with the Boost library aka 100 error lines of TEMPLATE EXPANSION TEMPLATE EXPANSION TEMPLATE (you missed a comma) TEMPLATE EXPANSION...

>>58369335
>use maps, reduces, folds, ...
I think you meant "use folds, folds and folds".

>>58372027
> Visual Studio supports std::thread and std::mutex and C++14, yeah?
Yeah, but it's still far behind gcc and clang.
> minGW lacking every other thing
Why don't you use msys2 instead, it has gcc 6.2 and clang 3.9 in the repository.

>>58369407
There are literally 19,683 possible states.

>>58372027

MinGW-w64 supports C++14, std::thread, and std::mutex just fine. In fact, it kind of had C++14 features a lot earlier than Visual Studio. There is no need for Boost to use any of these. Stop smoking crack and making shit up, Anon.

>>58370841

https://g.sicp.me/books/

https://youtu.be/48kP_Ssg2eY
This has me in tears laughing. I'm not sure if it's Scott being funny or just me getting relief for the frustration I have.
But it feels so good.

>>58371890
Indeed. A binary tree and a linked list is (can be) basically equivalent. Instead of for example Left and Right you have Next and Data. Same shit. It's just a matter of perspective. (Unless of course, by data you mean 3 different pointers or something.)

>>58372027
Maybe you didnt set your standard version?
When you use msvc you use it for the visual studio debugger or because you have no other choice.

>>58372027
>>58372109
Other MinGW distros also have 6.x.

Worst aspects of Haskell (implying GHC)?

>lazy evaluation
>lazy IO (God help us)
>GHC error messages (simultaneously completely unambiguous and totally unintelligible)
>type String = [Char]
>nearly typeclass-less Prelude functions
>laughable number of partial Prelude functions
>monads from 'return' and 'bind'
>exception handling, and the complete lack of consensus on the same
>record field name shenanigans
>{-# IHaveMoreExtensionsThanSLOC #-}
> :: and : are quite simply the wrong way around

That was just off the top of my head.

>>58370841
I just use bookzz.
>>58369407
Have a function that checks horizontal vertical and diagonal win conditions.
Have the checks wrap the edges of the play field. I.e diagonal (45 degree) check is arr[x][y]==arr[x-1][y-1]==arr[x+1][y+1]
Just add a %boardsize (3 in this example) within those brackets and you wrap the board. So when you check X=0 Y=2 it just werks.
You can also make it using a for loop for each condition to handle arbitrary size boards easier.

That way your program can just
-ask for input
-pass the input to the win check function
-if you've won you tell the player
-if not ask the other player to place a marker

I'd write an example but I'm on my phone.

>>58372613
You can of course generalise it to check arbitrary lines along arbitrary angles.

Then you can try and make special rules to make it a solved game still.

Would be far more interesting. Could probably make it an app and make millions. That 1024 game managed to make addition into a game. Why not redesign tick tack toe some?

File: questionmark.gif (23KB, 140x171px) Image search: [Google]

23KB, 140x171px

Not really prog but I need help

I want my website to play a sound every time a click is performed but I don't know shit about JS

File: FP.png (190KB, 2440x1228px) Image search: [Google]

190KB, 2440x1228px

>>58372548

>>58372548
>lazy evaluation
>lazy IO (God help us)
>type String = [Char]
>nearly typeclass-less Prelude functions
>laughable number of partial Prelude functions
>record field name shenanigans
>exception handling, and the complete lack of consensus on the same
These are bad.
>monads from 'return' and 'bind'
I don't get this.
> >{-# IHaveMoreExtensionsThanSLOC #-}
So, why don't they enable them by default? Why don't they make a new standard already? It's not like there's another competing implementation that needs to catch up.
> :: and : are quite simply the wrong way around
They are exactly right.

>>58372548
Prelude, rebindable syntax isn't flexible enough, type classes aren't flexible enough, lack of dependent types and type level lambdas and partial type application

>>58372683
>So, why don't they enable them by default? Why don't they make a new standard already? It's not like there's another competing implementation that needs to catch up.
The Haskell 2020 committee has been formed for this purpose, but Haskell 2010 was perhaps too conservative.

One of the things about the new extensions is the loss of principal typing - i.e. the idea that any expression in the language can have its type inferred.
(Rather than just have the user declare the type, and typecheck it)
Pretty much every extension breaks this in some regard, but since there are so many in standard usage, and since most types are still inferrable anyway, I hope a lot of these make it in.

Really interestingly would be if the committee adopted proposals for features that weren't even in GHC.

>>58372681
If you can't into zygohistomorphic prepromorphisms, you shouldn't be contributing to this discussion.

>>58372683
>monads from 'return' and 'bind'
Should be 'unit' and 'join'.
>{-# IHaveMoreExtensionsThanSLOC #-}
They're pissing about too long before coming up with the new standard. The reason they're chickening so hard is essentially this: >>58372718
>They are exactly right.
If you're typing out ':' more than '::', you're doing two things wrong:
1) Using too much explicit recursion
2) Not documenting your code enough

>>58372687
Aha, it's RebindableSyntaxbro. I'm sure you'd be much happier over here:

http://www.idris-lang.org/

>>58372674
I suggest you learn to search. What would you search for to figure this out?

>>58372760
>http://www.idris-lang.org/

Is the website written in Idris? It's remarkably slow.

>>58372760
>I'm sure you'd be much happier over here:
Nope.

>ridiculous excesses of language specialised syntax that needn't be
>"just do everything with syntax"

>>58372674

<audio id="audio" src="http://www.soundjay.com/button/beep-07.wav" autostart="false" ></audio>
    <a onclick="PlaySound()"> Play</a>
    <script>
    function PlaySound() {
          var sound = document.getElementById("audio");
          sound.play()
      }
    </script>

>>58372835
Yeah, I don't know what Edwin's obsession with syntax is. Puts me right off.

Plus the whole lack of type inference. I know it's undecidable, but I'm lazy as fuck.

That said, if you're this into type level stuff, you might want to a consider a 'real' dependently-typed language like Idris (and friends), rather than just Haskell With Gratuitous Extensions.

>>58372818
It's WordPress, ahuehue.

>>58372890
I want Idris but Haskell at the same time

>>58372913
Me too, bro. Me too :(

Idea for a videogame mechanic in a detective-style game:

The player (detective) is given a blurry security camera photo of the criminal. He has to describe the criminal to a computer using sliders for various qualities (nose width from skinny to wide, nose length from flat to long, ear size, skin color, etc.) and the game randomly generates a dozen or so possible criminals and then puts them on a graph with one of them being the actual criminal (not randomly generated). The game then uses the player's description with a k-nearest-neighbor algorithm and gets you a list of 10-15 criminals. If the player did well at describing the criminal, he shouldn't have to question too many dead end suspects before he gets to the real criminal. If he did poorly, he may have to try describing the criminal again.

File: NowItPlaysMusic.png (757KB, 856x576px) Image search: [Google]

757KB, 856x576px

Managed to get my Python bot for Discord to play music.

Took a bit over 200 new lines of code and maybe 400-500 lines if including debugging

>>58372984
I'll be taking a nap now, though. Just came by to share.

>>58372979
Sounds good anon but for no-dev game idea discussion you will find >>>/vg/agdg/ to be more receptive.

>>58373006
sleep tight

>>58372979
>>>/v/

>>58373019
People who can't dev don't tend to mention the k-nearest-neighbours algorithm.

That said, I could be wrong.

File: 1481407139437.jpg (67KB, 680x907px) Image search: [Google]

67KB, 680x907px

FUCKING C++ All this just to get a fucking path to module REEEEEEEEEE

    wchar_t pszModulePath[MAX_PATH];
    wchar_t pszModuleDrive[_MAX_DRIVE];
    wchar_t pszModuleDirectory[_MAX_DIR];
    wchar_t pszModuleName[_MAX_FNAME];
    wchar_t pszModuleExtension[_MAX_EXT];

    if (GetModuleFileName(NULL, pszModulePath, MAX_PATH) == 0) {
        DWORD dwLastError = GetLastError();
        PRINT_MSG(L"Failed to retrieve module filename.");
        PRINT_ERR_CODE(dwLastError);
        return;
    }

    errno_t error = _wsplitpath_s(pszModulePath,
        pszModuleDrive, _MAX_DRIVE,
        pszModuleDirectory, _MAX_DIR,
        pszModuleName, _MAX_FNAME,
        pszModuleExtension, _MAX_EXT);

    if (error == EINVAL || error == ERANGE) {
        DWORD dwLastError = GetLastError();
        PRINT_MSG(L"Failed to retrieve module filename.");
        PRINT_ERR_CODE(dwLastError);
        return;
    }

>>58373029
>>58373019
Thanks. I don't actually do any game dev because I lack creativity but I thought it would be a cool gimmick

>>58373089
Meant module name and drive

>>58373060
KNN is baby tier are you serious.

Get this crap out of here.

>>58373060
I posted the original comment; I do dev all the time. I don't even do game dev ever because I can't make art. Here's a knn I implemented:

https://github.com/collinoswalt/k-nearest-neighbor/blob/master/knn1.c

figures out if your listing is a house or apartment with training data from my area (Northern VA)

>>58373089
How is that the fault of C++? It's just WinAPI being shit.

>>58373102
Yeah, but most people who present video game ideas tend to know literally no CS: consider 'hey guise what about this game that is like Eve but with more GUNS lawl'.

And here he's been kind enough to point out that he does in fact, know some CS (as we surmised earlier): >>58373120

>>58373120
>he didn't implement kd-trees

>>58373060
No-dev in the context of agdg means you're a procrastinator. Not that you don't know how to develop software.
Also don't try to force your comp-sci into games. There's better solutions to the problems you're facing in games.

>>58373153
I'll look into that for the future, thanks!

>>58373141
If you want to attract even more /agdg/fags here then be my guest. Sure let's all just field our lame idea-guying ITT.

noob here

is using recursion instead of a loop pretentious?

>>58373169
>>58373175
I'm vindicated by the correctness of my deduction that he did, in fact, know some dev, so further discussion isn't really relevant.

>>58373189
Write a function to insert into a binary tree in both a recursive manner and using loops and you'll get your answer.

>>58373189
recursion a shit

>>58373169
>Also don't try to force your comp-sci into games. There's better solutions to the problems you're facing in games.
Not anon, but surely there's a reasonable overlapping of software design for games and non-gaming programming? Or do you mean "comp-sci" as in making everything a template/only using OO Design Patterns?

>>58373172
They are a bitch, gl.

>>58373189
Recursive versions of functions is sometimes easier to implement because you have the stack managed for you implicitly.
To arbitrarily favor recursion would be pretentious.
But likewise writing everything in a loop form is usually a waste of time. There's situations where you should make your recursive algorithm a loop instead but let what is easier to write determine how your draft looks.

Tdlr as with anything software write what's easiest to write and then fix your shit if needed.

>>58372400
And people seriously try to argue that C++ is good.

snekmeme noobie here, how would i go about writing a function for a irc bot that takes random parts of sentences of users, stores them and says a random concatenation of them given a small time period?

>>58373308
google markov chains

>>58373308
import markov_chains

>>58373259
>overlapping
It's there. But you can't guide yourself from the data structure and algorithms approach. Games have so much fuzzy code in it that just not guided by any higher level reasoning that you shouldn't put too much effort into thinking about the CS side of things.
The CS side of things can be useful for everything below your engine API. Like how you handle draw commands and how you lay out your world. But if we take anons example you may find that a KNN tree doesn't produce the similarities humans perceive as relevant even if you match the suspects well by that standard. So you have to do something else.

>>58373318
>>58373350

thanks dudes

File: 1483423464145.jpg (96KB, 576x720px) Image search: [Google]

96KB, 576x720px

>>58373350

import a_function_for_a_irc_bot_that_takes_random_parts_of_sentences_of_users_stores_them_and_says_a_random_concatenation_of_them_given_a_small_time_period

Fucken Python.

>>58372400
Just finished watching.

I didn't know C++ was this much of a mess.

Reminder:

Idris > Haskell > Scala > OCaml > F# > ATS > C# > C++ > D > Java > C > Rust > Fortran > Common Lisp = Scheme = Clojure > Forth > Perl > Python = Ruby = JavaScript = PHP > Go

Yeah, Ocaml really is the lang.

>>58373385

from __future__ import my_irc_bot

Haskell is dead

https://www.reddit.com/r/haskell/comments/5m678s/suggestion_code_of_conduct_for_haskell_community/

>>58373406
> C# > C++ > D > Java > C > Rust
You fucked up this part, m8, it actually goes like this:
Rust > C# > C++ > C > Java > D

>>58373459
C++ and Java are too high, m8.

>>58373305
>>58373399
And this is just the surface of it really.

File: 1483730818833.png (160KB, 373x345px) Image search: [Google]

160KB, 373x345px

>>58373445
Much more Pythonic, thank you.

>>58373457
Kill yourself

File: toj.png (12KB, 246x200px) Image search: [Google]

12KB, 246x200px

https://users.rust-lang.org/t/how-do-we-combat-the-apparent-culture-of-racism-and-sexism-in-rust/8754/14

>>58373551
The nose has spoken.

>>58373551
>>58373584
The nose knows

>>58373551
> it's another episode of "woke hacker 4chan tries to provoke stupid SJWs but instead gets weighted and rational answers"
It's as if he's the only one who cares about the CoC at all, the rest being busy actually writing code.

Just started experimenting with Go.
Is it possible to use the multiple return values of a function as arguments for a function that has additional arguments on top of that?

Example:

func multipleReturn() (int, int) {
    return 1, 2
}

func f1(x int, y int) int {
    return x + y
}

func f2(x int, y int, z int) int {
    return x + y + z
}

func main() {
    // Works
    f1(multipleReturn())

    // ERROR: multiple-value multipleReturn() in single-value context
    // not enough parameters in call to f2
    f2(multipleReturn(), 3)
}

where is the fpt thread?

>>58373635
But I didn't post in that thread at all. I just came across it. Carol might actually make me switch over to D, though. If I have to hear how her middle-class white-ass is a "underrepresented minority" I'm going to puke.

>>58373646
In Python that's just:

def multi_ret():
    return 1, 2

def f1(x, y):
    return x + y

def f2(x, y, z):
    return x + y + z

print(f2(*multi_ret(), 3))

>>58373646
I doubt it, because Go has decided to pretend its tuples aren't really tuples because that would be too FP oriented, and instead decided they're "multiple return values" (and so cannot be used as tuples)

>>58373646
nope

Working on a new way of compressing images: filtering * any lossless - gain 10 to 20 percents:
http://pi-xel.io

>>58373646

Why do you want to do that?

Pass a pointer to a callback, run that in a function using a channel. That's why they made Golang, easier "concurrency"

File: NOwS64t.png (18KB, 600x600px) Image search: [Google]

18KB, 600x600px

>>58373677
> Letting other peoples' political believes influence your technical decisions

>>58373795
My health is more important than a programming language.

C > C++

>>58373870
quality post my man

>>58368865
Why is Shinji dressed like a girl?

Guys I need a good capstone idea. I have no imagination. should I do a ANN or maybe something with data analyst, like a recommendation engine, maybe for music similar artists like a mobile app for it?

>>58373906
Shinji is a girl at heart

>>58373906
it's Maya

>>58373923
Instant messaging with full public key crypto and groups, such that the content of messages cannot be discovered by anybody other than the participants

>>58373946
thanks I like that idea a lot actually

maybe an ANN to play checkers would be a cool idea too

>>58373884
Thanks bro, if you like what I write please support me on patreon

>>58373964
ANN to recognize different anime characters, then make it into a public API and write a Firefox extension so I can hover over an image on 4chan to get the name and the series the character is from.

>>58374024
You could just write a Firefox extension that reverse image searches and returns Google's 'best guess for this image' text.

That way you wouldn't have to write a neural net in JavaScript.

>>58373662
This is ftp thread home away from home right now, since fpt is very inactive except during holidays, since most fpters actually have jobs.

>>58372850
>>58372674
Thanks but this is not exactly what I was looking for

I dont want the sound to play only when I click on an object, I want the sound to play when I click anywhere on my website

This script is only playing the sound when I click on Play

>>58374132

document.body.addEventListener('click', function() {
    document.getElementById('audio').play();
});

>>58374024
I kek'd

>>58374024
this

File: 2017-01-07-173821_1600x900_scrot.png (124KB, 550x577px) Image search: [Google]

124KB, 550x577px

>>58374024
>>58374060
>>58374171
>>58374210
Chrom* has this functionality built in.

Let's say I have this string

"Unique_Identifier_1","RandomNumbers1":1483432202,"RandomNumbers2":"27.9"
"Unique_Identifier_2","RandomNumbers1":1483432202,"RandomNumbers2":"27.9"

How do I get the 27.9 from

Unique_Identifier_1

out of this with regex?

I'm not sure what to do since the randomnumbers1 is always inbetween the content I want.

>>58374324

"Unique_Identifier_1","RandomNumbers1":[0-9]+,"RandomNumbers2":"[0-9]+"

No escape sequences.

>>58374339
Forgot the decimal point in the 27.9, my bad.

>>58374357
>>58374339
Thanks family, I got it.

>>58374324
I think what ur looking for is called "capturing groups"

In C++, is an array of pointers to keep track of relevant objects a bad idea?

>>58374597
>c++
>bad idea
YES

>>58374612
What's the matter brainlet, afraid to deal with some complexity?

is there anything javascript can't do?

>>58374643
Not be shit

>>58374643
be a good programming language

>>58374643
Be well designed

>>58374597
yes, check out "data structures" my dude. it's really cool

>>58374655
>>58374656
>>58374661
>do
>be

>>58374597
Why not just a vector of objects?

>>58374666
Sorry Satan but I haven't taken that course yet.

Now what the fuck do I do if my project requires me to do something that's beyond my scope.

>>58374707
give up

I want to make a Java scraper that would need to log in in order to scrape.

Until now I've only had to make easy scrapers that just sent a POST/GET request and parsed the HTML.

How are logins usually handled with manual POST/GET requests?

>>58374697
I had some reasons not to use the STL but I may.
Or maybe I could allocate an array with malloc and write a template to add elements to the array.

>>58374726
Pussy.

File: browser_.webm (3MB, 862x728px) Image search: [Google]

3MB, 862x728px

>when you code in your dreams

>>58374748
>sepples
>malloc
you are doing it wrong

How the fuck was clojure designed? Did some guys got drunk and said "Let's ignore all lisp conventions in our lisp dialect?"

For example why the fuck are they using def and defn instead of define (like racket) or at least defun? Especially since it's lisp-1... Or calling hashes/dictionaries maps... This is just going to cause confusion with map function.

Also using ((or + -) 2 3) as a legit example, while it's just very bad style?

I so hope I am wrong, I just began studying it, but it seems retarded. Is clojure python of lisps?

>>58374760
that looks like visual cancer

>>58374692
What?

>>58374840
Or making keywords their own datatype, which is probably the redardest decision of all.

>>58374840
clojure is not lisp.

>>58374849
yes (sorry for bad english)

>>58374859
>>58374863

I would like to agree with you, but they are literally advertising it as lisp. It's horrible.

Also using fn instead of lambda for anon functions... what were they thinking? And not implementing tail call optimization... in lisp-1 at that.

Is clojure some retarded cousin of python?

>>58374917
>And not implementing tail call optimization
It's trying to be Lisp that runs on the JVM. Hence why it has no TCO.

>>58374840
>>58374859
Clojure is a lisp designed for non-lispers, with total disregard to lisp's 50 years of idiosyncratic tradition. This is why it's the best lisp out there.

Noob question here, but how do I start writing web crawlers of some sort? Or more simply, a program that can access a webpage, and return some value there?

For example, suppose I want to write a program that goes to weather.com and tells me the local weather (taking zip code as input, say). What's the best language to write this in, and where can I get started?

>>58369704
show a screencap of your wallet (black out addresses)

Holy shit just got a second monitor and a monitor stand, got one horizontal and one vertical, why did no-one tell me programming could be so comfy?

>>58375001
Do an Ajax request

>>58374951
If I understand correctly it compiles to jvm bytecode, right?

Does jvm bytecode lack go-tos or other iteration constructs?

>>58374998
Throwing away all progress of past 50 years, all the culture that was established around it and at the same time failing to have any sort of internal consistency is a good thing?

>>58375071
> Throwing away all progress of past 50 years, all the culture that was established around it [..] is a good thing?
Yes, if it is a dead end.

>>58375071
>If I understand correctly it compiles to jvm bytecode, right?
That's the only way to run it on the JVM, so yes.
>Does jvm bytecode lack go-tos or other iteration constructs?
It has the 'goto' instruction. It's just that the JVM isn't meant for FP languages.

>>58375106
Tail cail optimization doesn't have anything to do with FP. Even fucking assembly has recursion.

Tail cail optimization should just replace recursive function call with gotos (well not just gotos, but basically)... I don't see what is the problem here.

>>58375103
Show me how the fuck are common lisp and racket dead ends.

>>58375064
Is there anything for C++? Or something other than Javascript?

>>58375010
Notice how there is no response. Now stop replying to this obvious troll.

>>58375167
Well write a javascript code generator, it's trivial (for simple things) really in non-retarded languages.

>>58375167
I dont know.
Ive heard Python is good for writing web crawlers.

>>58375167
libcurl

>>58375186
The point is I don't know any javascript.

File: dead_end.png (37KB, 1163x599px) Image search: [Google]

37KB, 1163x599px

>>58375154
> Show me how the fuck are common lisp and racket dead ends.
CL is stagnating for the last 20 years, it's full of obsolete and idiosyncratic ideas no one cares anymore, it's too hard to jump into and it just isn't worth it anymore, other languages evolving so fast at the same time. I have nothing bad to say about Racket, I actually like it.

>>58374863

Now I see... Clojure isn't lisp, its vecp. Vector processing language.

new Thread("anime.jpg");

when

New thread:
>>58375309
>>58375309
>>58375309

>>58375231
Thanks, this helps a lot. I tried curl in the terminal and it's a start.

Could anyone say something about security when using curl and/or wget? I'm going to learn about these, but I care a lot about encryption and wonder what I have to do to send encrypted requests/verify domains/etc.