/dpt/ - Daily Programming Thread

https://danluu.com/programming-books/ :)

If you want to talk more in real-time, we also have a new Discord Server:
https://discord.gg/69Bc6X

File: JAPCurly.png (183KB, 373x489px) Image search: [Google]

183KB, 373x489px

I love Curly Braces

>>62227640
>he fell for the programming meme

>>62227640
https://en.wikipedia.org/wiki/Category:Lisp_people

Wew, it's like a Tel Aviv phonebook.
Why does (((Lisp))) attract so many of (((them?)))
Is there EVER an excuse for goyim to use this subversive language?

>>62227758
These jews are friends who we can truly trust.
They would never make this programming work for free otherwise.
https://clisp.sourceforge.io/

>>62227788
>These jews are friends who we can truly trust.
t. Abraham.

why can't I route my packets the way I want to

>>62227758
((They)) are over represented in academia
Lisp is an academia language
Hmm

cmake_minimum_required(VERSION 3.0.2)

foreach(i RANGE 1 100)
    set(should_print_int TRUE)
    math(EXPR mod3 "${i} % 3")
    math(EXPR mod5 "${i} % 5")
    if((mod3 EQUAL 0) AND (mod5 EQUAL 0))
        message("FizzBuzz")
    elseif(mod3 EQUAL 0)
        message("Fizz")
    elseif(mod5 EQUAL 0)
        message("Buzz")
    else()
        message("${i}")
    endif()
endforeach(i)

JUST
FUCK
MY
SHIT
UP
I'm sorry to make you all have to look at this

import random


def diceroller(d):
    wrap = input("Wrap? ")
    quantity = int(input("How many times? "))
    lengthcheck = True
    resultlist = []
    wrseultslist = []
    while lengthcheck:
        if wrap == "no":
            resultlist += ([random.randrange(0, d + 1)])
            if len(resultlist) == quantity:
                lengthcheck = False
        else:
            wrseultslist += ([random.randrange(d / 2 + 1, d + 1)])
            if len(wrseultslist) == quantity:
                lengthcheck = False
    while wrap == "yes":
        print("Wrapped Result:", wrseultslist)
        break
    while wrap == "no":
        print("Result:", resultlist)
        break


sides = int(input("Roll a D"))
print(diceroller(sides))
input("press enter to close")

How the FUCK does fmtlib work?
I've tried many times to implement compile-time type safe formatted printing supporting user defined types in C++, but every time I try I cannot find a way to do it.
Is fmtlib magic?

>>62228045
why even use ifs if we can use a while with a break at the end

why even live

>>62227640
rewriting my hello world program in haskell

File: 1421164417027.png (342KB, 714x738px) Image search: [Google]

342KB, 714x738px

Employed Haskell programmer reporting back

>>62228610
> <programming language> programmer
this is so stupid
you can learn haskel in a week and suddenly you are 'a haskel' programmer

you don't specialize by language in programming

File: doubt.png (31KB, 300x162px) Image search: [Google]

31KB, 300x162px

>>62228643

>you can learn haskell in a week

Languages that will get you jobs easily:

>javaScript
>Java
>c#
>c/++
>clojure
>SQL

Languages that might get you jobs:
>ruby
>python
>swift
>go

Languages that won't get you jobs:
Everything else

>>62228679
It's easy to find a job as an expert MUMPS programmer, over 100k too.

GREPTHIS()
       NEW SET,NEW,THEN,IF,KILL,QUIT SET IF="KILL",SET="11",KILL="l1",QUIT="RETURN",THEN="KILL"
       IF IF=THEN DO THEN
       QUIT:$QUIT QUIT QUIT ; (quit)
THEN  IF IF,SET&KILL SET SET=SET+KILL QUIT

>>62228679
How do I learn sql?

>>62228679
What about mongo?

>>62228737
https://www.youtube.com/playlist?list=PL32BC9C878BA72085

>>62228757
Mongo isn't a language it's just json shit anon

>>62228737
Select sqltutorials from Google
>>62228757
Used by a few big companies but mostly irrelevant otherwise. A lot of virtual private server providers have added support for mongodb which might help with the demand.

>>62228770
And sql is mostly English. Your point? It's the additional functions like join in sql that matter.

File: [email protected] (12KB, 1224x618px) Image search: [Google]

12KB, 1224x618px

>>62228679
>tfw you're learning python right now

b-but you guys told me C# and Java are languages that get outsourced to pajeets.

>>62228679
really clojure?

File: 1503781261192.jpg (1MB, 620x4462px) Image search: [Google]

1MB, 620x4462px

It goes like this.

>>62228800
It makes for a good base language and a lot of the bigger companies like Google are using python. No harm learning it. You should go deeper by learning C though.

>>62228679
if this is true how come there are so many employed haskell devs in here?

>>62228801
Clojure is extremely popular with defense contract jobs for AI, and since self driving cars are becoming a big deal there are more jobs than people who can write it professionally

This will likely fall as time goes on though. The reason it's so employable right now is because so few people were doing it

>>62228848
>Clojure is extremely popular with defense contract jobs for AI
lol, source?

>>62228737
The PostgreSQL manuals are really good.
https://www.postgresql.org/docs/9.6/static/index.html

>>62228848
>Clojure is extremely with defense contract jobs for AI
stop lying

>>62228858
He's a shitposter, stop giving him (you)s

File: tr-being.jpg (152KB, 1000x1000px) Image search: [Google]

152KB, 1000x1000px

Java is the most intelligent language.

>>62227758
because (((lisp))) uses the parentheses a lot.

>>62228892
Don't bully java

How harmful would it be to reinterpret_cast a

void (*)(const T*)

to a

void (*)(const void*)

?

>>62228948
That should be well defined.

>>62228948
>>62228964
oh wait nevermind that may be a bad idea

>>62228978
Can you explain?
I am making sure that the object being pointed to is correctly aligned.
I don't really have any other choice.

I know that casting a T* to a void* is well defined, so I think >>62228948 should be fine, but I'm not sure.

>>62227640
How does Tkinter (the Python module) work? I compiled Python 3.7 and ldd'ed the resulting binary, and they practically nonexistant. I assume the only way Python can make something graphical is to:

1. Fire up dynamic linker library.
2. *Try* to find tkinter library.
3. Proceed on success and fail if not found.
4. All thins probably happens on import tkinter line.

Is this correct? Is there any way to force Python binary to link with tkinter at compile time? I mean, there's a tool "freeze", which produces a single executable from your python program, but that's not quite correct if it doesn't hardwire tkinter into it. So if you make a GUI in Python and use freeze to emit an executable, it will crash if the user doesn't have tkinter installed. So you have to ship tkinter.so or tkinter.dll together with the app. Is that correct?

>>62229013
Pointer types may be of different sizes. void* is guaranteed to be the size of the greatest of all those, like a union.
It's not guaranteed that T* is equal to the size of void* so that might screw up the call according to your platform's calling conventions. A T* can always convert to a void* but it isn't already a void*.
In practice it may work (I'm not sure what platforms use different sized data pointers) but I don't think it's guaranteed safe.

How's this going to help you with alignment?

>>62228892
lmao

>>62227640
Anyone got a link to an up-to-date version of Haskell Book? Too cheap to pay $60 for a PDF

>>62229251
Just read this: http://learnyouahaskell.com/

>>62229440
>learnyouahaskell
how cringe-inducing and puerile

>>62227758
>jews are successful at engineering
wew lad

>>62228679
>clojure
>anything but NEET-lang
lol

>>62228812
>t. 13 year old who read a bunch of tutorials

>>62229477
This. I fucking hate haskell devs so much
>learn you a haskell m'good sire!
Neck kysself

File: jhvanigd.png (138KB, 396x385px) Image search: [Google]

138KB, 396x385px

>contact for our client is a "software engineer" like me
>start discussing our suggested approach for him
>can feel how everything just passes right over his head
>turns out he doesn't have an MSc and only has a BSc in CS or IT or something

>>62229073
Ok, that's weird.
Anyway, I think I found a way around it. I'm implementing type-safe formatted printing that supports user defined types (although I couldn't find a way to do it entirely at compile-time).

#include <iostream>
#include <type_traits>
#include <stdexcept>

namespace detail {
    template <typename T>
    static void print(std::ostream &o, const void *t) {
        o << *static_cast<const std::remove_reference_t<T>*>(t); // Probably the most questionable line in this code, but if static_cast allows it, it must be fine, right?
    }
    
    using Fp = void (*)(std::ostream &o, const void*);
}

template <size_t len, typename ...Args>
std::ostream &print(std::ostream &o, const char (&fmt)[len], const Args& ...args) {
    detail::Fp fp[] = {detail::print<Args>...};
    const void *ap[] = {&args...};
    
    size_t arg = 0;
    
    for (size_t i = 0; i < len; i++) {
        if (fmt[i] == '{') {
            i++;
            
            if (arg >= sizeof...(args)) {
                throw std::out_of_range("print(): Too many format specifiers");
            }
            
            if (fmt[i] != '}') {
                switch (fmt[i]) {
                    case 'x':
                        o << std::hex;
                        break;
                }
                
                i++;
            }
            
            fp[arg](o, ap[arg]);
            o << std::dec;
            arg++;
        }
        else {
            o << fmt[i];
        }
    }
    
    return o;
}

template <size_t len, typename ...Args>
std::ostream &print(const char (&fmt)[len], const Args& ...args) {
    return print(std::cout, fmt, args...);
}

template <size_t len, typename ...Args>
std::ostream &println(const char (&fmt)[len], const Args& ...args) {
    return print(fmt, args...) << '\n';
}

struct Vec2 {
    float x, y;
};

static std::ostream &operator<<(std::ostream &o, Vec2 v) {
    return o << '(' << v.x << ", " << v.y << ')';
}

int main() {
    int x = 57;
    
    println("x = {}", x);
    println("{}", Vec2{3.6, 12});
    println("test {}", "asdf");
    println("0x{x}", 4096);
}

$ g++ main.cxx 
$ ./a.out 
x = 57
(3.6, 12)
test asdf
0x1000

As far as I can tell, there's no UB in this code, but maybe someone here knows better than I do.

>>62229551
degreelets should not be allowed to work on complex softwares
they'll just cause trouble

File: K85ZWV2F_400x400.png (108KB, 400x400px) Image search: [Google]

108KB, 400x400px

Can we agree that Java is the best language to learn for beginners?

It's simple on the surface, yet powerful and complex when you get further into it. Easy to make simple programs, and immediately provides newcomers an opportunity to understand object oriented programming, which is so widely used in the industry today.

>>62229792
>powerful

uh not really

>complex

since when is this a good thing?

>>62229792
>Can we agree that Java is the best language to learn for beginners to OOP
Change it to this and you'll be right.

>>62229550
Sounds like you've never talked to a single one.

just learned haskell in 20 minutes

>>62229879
still learning after 6 years, ask me anything

Are the courses on CodeAcademy, Udemy, Linda and Treehouse worth the money?

is the 2017 non touch bar macbook good for programming?

>>62229890
how do I do a for loop

What condition do I have to meet to stop the loop of me coming to 4chan and shitposting every day?

>>62229943
True

>>62229947
False

>>62229892
>paying for information on the internet
no

File: 972853.jpg (125KB, 1280x960px) Image search: [Google]

125KB, 1280x960px

Why is dpt so autistic?

>>62229909

for_ [1 .. 10] $ \i ->
  print i

>>62230019
We program.

>>62230005
All the things schools teach kids is available on the internet; are you saying schools are obsolete?

>>62230033
no, if you go to school (and finish it) you'll get a certificate that actually means something
on the internet you don't

File: 1480529706908.jpg (16KB, 251x242px) Image search: [Google]

16KB, 251x242px

>>62230029
>We program.
Good one, m8

How do you monitor your remote daemons?
Simply Unix system email on errors?
Some fancy logging daemon?

>>62227640
Stopwatch in 2 hours and 50 lines of code, upgrade from my previous one for CLI

#!/usr/bin/python3

import tkinter

"""Global variables."""
root = tkinter.Tk()
stopwatch = tkinter.StringVar(root, "00:00:00")
mouseGrab = (0, 0)  # mouse coords on last left button press (window-wise)


def mouseLeftButtonPressed(e):
    """Called when left mouse button is pressed."""
    global mouseGrab
    mouseGrab = e.x, e.y


def mouseLeftButtonDragged(e):
    """Called when left mouse button is dragged."""
    x, y = e.x_root - mouseGrab[0], e.y_root - mouseGrab[1]
    root.geometry("+{}+{}".format(x, y))


def mouseRightButtonPressed(e):
    """Called when right mouse button is pressed."""
    root.destroy()


def update():
    """Called once right before the main loop."""
    h, m, s = [int(x) for x in stopwatch.get().split(':')]
    n = h * 3600 + m * 60 + s + 1
    h, m, s = n // 3600, n % 3600 // 60, n % 3600 % 60
    stopwatch.set("{:02}:{:02}:{:02}".format(h, m, s))
    root.after(1000, update)

def main():
    """The main function."""
    root.overrideredirect(True)
    label = tkinter.Label(root, textvariable=stopwatch)
    label.bind("<Button-1>", mouseLeftButtonPressed)
    label.bind("<B1-Motion>", mouseLeftButtonDragged)
    label.bind("<Button-3>", mouseRightButtonPressed)
    label.pack()
    update()
    root.mainloop()


main()

Tested and works fine on Lubuntu

File: Untitled.png (204KB, 1891x767px) Image search: [Google]

204KB, 1891x767px

yo, /dpt/. I'm trying to get clover to build properly on my machine, but no matter how many times i follow the steps on the wiki page for it, i keep getting 'gradle project sync failed' in the top of my editor window, and 'createprocess error=2, system cant find file specified' in the messages window. i have the current version from floens's own repo.
The wiki mentions to get the correct SDK tools and SDK platforms from the SDK manager, and i ahve them.
the platform value I have is API level 25, and as you can see the sdk tools value also matches what is suggested in build.gradle. since that's the only step in the wiki that is outside the project, i'm confused as to why gradle isn't finding this missing file. Does someone know what I can do to fix this?

>>62230135
The mouse doesn't work for me.

>>62230135
Now you can talk to God

>>62230167
Works on my machine.

>>62230057
and you're saying people should get certificates for gathering info from the internet. Genius.

>>62230283
'Should' in what sense? For your own personal gain, yes, you should. By any other metric, probably not.

Thoughts?

public class checkGUI extends JFrame {
    private JTextField textfield;
    private JCheckBox checkboxBold;
    private JCheckBox checkboxItalic;
    
    public checkGUI(){
        super("checkerbox");
        setLayout(new FlowLayout());
        
        textfield = new JTextField("Enter sentence", 20);
        textfield.setFont(new Font("Serif", Font.PLAIN, 15));
        add(textfield);
        
        checkboxBold = new JCheckBox("bold");
        add(checkboxBold);
        
        checkboxItalic = new JCheckBox("italic");
        add(checkboxItalic);
        
        HandlerClass handler = new HandlerClass();
        checkboxBold.addItemListener(handler);
        checkboxItalic.addItemListener(handler);
    }
    
    private class HandlerClass implements ItemListener{
        public void itemStateChanged(ItemEvent event){
            Font font = null;
            
            if(checkboxBold.isSelected() && checkboxItalic.isSelected())
                font = new Font("Serif", Font.BOLD + Font.ITALIC, 19);
                
            else if(checkboxBold.isSelected())
                    font = new Font("Serif", Font.BOLD, 15);
            
            else if(checkboxItalic.isSelected())
                font = new Font("Serif", Font.ITALIC, 15);
            
            else
                font = new Font("Serif", Font.PLAIN, 15);
            
            textfield.setFont(font);
        }
    }
}

>>62230283
when did I imply that even in the slightest?

>>62230302
>>62230310
You literally implied schools are useless for anything but certificates because anything you can learn in school is available on the internet.

Embarrassing!

Discarded paid courses here >>62230005, meanwhile say certificates are the only useful thing schools offer >>62230057

Just stupid.

>>62230341
>reddit spacing

Speaking about certs, are Oracle or Microsoft certificates worth anything?

>>62228800
The big Ai-developing companies still need local boys tho.

>>62230341
>meanwhile say certificates are the only useful thing schools offer
yes, the certificate proves that you passed their tests, and have some sort of knowledge, without that you knowledge means nothing for an employer
if you are after knowledge for knowledge (which is a healthy and noble goal) you are good with the near unlimited free knowledge on the internet

also>>62230416

File: george-soros-europes-nightmare-is-getting-worse-and-only-germany-can-make-it-stop.jpg (476KB, 3107x2330px) Image search: [Google]

476KB, 3107x2330px

>>62227758
>Tel Aviv

File: 1451761096953.jpg (142KB, 404x547px) Image search: [Google]

142KB, 404x547px

>>62227697
She's a good girl

>>62227758
>tfw I am the only nonjew on my team
wew, feeling like a special snowflake tbqh

At what point does Swing become a liability, and something else start to look more appealing for a GUI in Java?
I'm looking to make a game for a uni project and we're discussing our options. People say Swing is slow, but how slow really, and when should I start caring about it?

>>62229551
i dont have a degree and am smarter than u

>>62230951
NetBeans is built on Swing
if you need an example for its speed and versatility

File: IMG_0880.jpg (92KB, 400x538px) Image search: [Google]

92KB, 400x538px

I been studying Python but my cousin wants a cookie clicker for iOS to show off to investors that they can pump out apps, and pitch some other shit.
So now I gotta learn my way around C# in visual studio with Xamarin. Still not sure if I'll have to do a 2D OpenGL / unity game or simple views.

>>62227640
>Let's study hard so we can succeed!
I could not finish that Haskell book. I get the idea behind it but it seems overly complicated. Couldn't even do a fizzbuzz or fibonacci in it. Takes like 5 seconds in any other language.

Playing with http4s. I'm still looking for a way to fold tasks into a single Task.

>>62231031
Eclipse is better than Netbeans?

>>62231139
Opinion

>>62231139
Functionally almost the same, both are good

>>62230951
You could try JavaFX. It's the successor of Swing.

>>62230951
is swing or javafx really viable for games? i've used libgdx and found it pretty nice to use, worth checking out

>>62231328
I don't know, and I don't know how to find out, there aren't really any answers online besides shallow "it bad"

You guys like to shit on Electron apps, but is there any other viable cross-platform gui toolkit out there for C/C++?

>>62231372
Qt

File: 1478198829016.png (250KB, 434x368px) Image search: [Google]

250KB, 434x368px

WHY THE FUCK IS IT TAKING SO LONG TO ADD ASSOCIATED TYPE CONSTRUCTORS TO RUST

>>62230951
What kind of game? And if you are using swing for a gui, what do you use for everything else? libgdx, lwjgl?

WebAssembly: oldschool desktop autists strike back (with the help of a weird rust crowd)

>>62231421
How would lwjgl be good for something other than the GUI when it's just an OpenGL wrapper API? Most of the code afaik would be written by team members themselves, we haven't decided what libraries we would use outside of swing or javafx for GUI.

Probably something real-time, top-down, I'm also looking to start a passion project in a similar vein that'd be an open world RPG kinda thing.

>>62231403
Because Rust.

>>62231403
stop using toy languages

>>62231523
This.

>>62231081

traverse

? Maybe

sequence

?

>C/C++ codebase
>I didn't write most of it
>build with Makefile

is there an easy/standard way to at least add compiler warnings which then can be turned on/off?

>>62231474
>How would lwjgl be good for something other than the GUI when it's just an OpenGL wrapper API?
I can't understand this sentence

Niggers, I need help
I am trying to execute INSERT INTO usuario (dni, pass) values (dni, pass);
The values fo dni and pass are C++ variables

set = sta->executeQuery("INSERT INTO usuario (dni, pass) values (dni, pass)");

It doesn't work

>>62231403
people still use that after seeing how metaprogramming in rust was even more ridiculous than c++?

File: 1494775377859.png (4KB, 183x275px) Image search: [Google]

4KB, 183x275px

>>62231594
>generics are metaprogramming

>>62231593
create a concatenated string or something and pass that

>>62231568
#define WARNING_ON
#idef WARNING_ON
#warning "text"
#endif

>>62231593
>>62231617
the way that works in c++ if you want your variables evaluated is
"sample text " + variable + " more text" +another_variable

>>62231593
Please, do yourself a favour and read up on sql injection and prepared statements.

How much do you love IBM's API?

>>62231533
I think traverse will do the job. Thanks anon.

>>62227640
Where is the best place to start for someone who has no programming or coding experience and has no access to formal education?

Asking for a friend.

>>62231739
Youtube

>>62231644
Like this?
set = sta->executeQuery("INSERT INTO usuario (dni, pass) values (" + dni + "," + pass +")");
It didn't work. main.cpp:70:76: error: invalid operands of types ‘const char*’ and ‘const char [2]’ to binary ‘operator+’
>>62231666
I didn't know those things existed, thanks for the information

File: geany.png (109KB, 3840x2080px) Image search: [Google]

109KB, 3840x2080px

>Not using geany

>>62231739
The book Head First Java.

>>62231782
>It didn't work. main.cpp:70:76: error: invalid operands of types ‘const char*’ and ‘const char [2]’ to binary ‘operator+’
then assemble a separate string and pass that

string s = "INSERT INTO usuario (dni, pass) values (" + dni + "," + pass +")");

don't forget to add the string header if you haven't already #include <string>
and you can explicitly convert shit to string with string() where necessary if it still complains

>>62231799
i was a heavy user of this back in 2008 or so.

- does it parse the entire project yet?
- does it have valgrind (memcheck and callgrind) support?

>>62231739
All you will ever need:

https://www.youtube.com/user/thenewboston/playlists

>>62232015
Wow, this is a lot of material. Thank you! Any suggestions on what playlist to start in? If not I will just dive in.

>>62232015
>thenewboston
why do you assume he's a brainlet like you are, brainlet

int values[9];
    if(something)
        values = {0, 20, 20, 30, 80, 70, 0, 0, 20};
    else
        values = {0, 40, 30, 50, 20, 20, 0, 0, 80};

how is this done in C exactly, wtf

>>62232152
https://godbolt.org/
Why don't you see for yourself?

>>62232152
C allows you to use uninitialized variables, so there's nothing special going on.

>>62232048
Java, then move to Python.

>>62232054
The only brainlet here is you.

>>62228392
Can't use if when it's in the same indent as other whiles

>>62232256
>>62232181
i asked because the code i posted doesn't compile

the question is, what is wrong, not how does the compiler do it

>>62232278
Yeah, you can actually

>>62228655
>>62228643
>>62228610
Ah cool. Do you mean that you write Haskell in production or you just a job and happen to write Haskell?

I'm picking up Elm and am learning functional programming myself. It feel pretty great ~

>>62229534
Pretty much any language can go extremely deep if you start working with it more and more. It just depends on how fast and maintainable you want to get shit done and what kind of performance you need. It also depends on what kind of audience you want to ship your code to.

>>62232508
i dont have a job nor do i know haskel
fp is fun yes, i read sicp i liked, even though ass holes will tell you scheme has nothing to do with fp

Trying to deal with x86 interrupts in long mode might be the most suicide inducing thing I've done yet. I already wasted my morning because I though I needed to use the pre-remapping address for lidt since that's what I did for lgdt, but it only worked since the gdt was set up before I disabled lower (identity) mapping.

Can someone explain to me the point of classes? Why would I ever create a class? Whats the benefit? If I only make small programs (say less than 300 lines), would there ever be a point in making a class?

>>62231985
>- does it have valgrind (memcheck and callgrind) support?
Why would you need this integrated?

>>62232856
>If I only make small programs
Keep doing what you're doing, neet, you haven nothing to worry about. Leave classes to the big boys.

>>62232953
>t. low wage code monkey
I get the point of classes for the standard library; it makes sense for them to expand the library with classes. But why would I ever use a class? It seems like it would lead to confusion and less readability for other users, who then have to decipher my class

Employed C programmers
never read k&r or modern approach

File: Triangulation.webm (574KB, 1024x768px) Image search: [Google]

574KB, 1024x768px

>>62227640
My triangulation algorithm had a bug in it, so I had to write this debugging visualization to figure out what the problem was.
The upside is that I now have a nice visualization of the ear-clipping algorithm.
Green = ear, blue=convex vertex, red=reflex vertex (interior angle > PI.)

It turned out that I carelessly made my GetVertexInteriorAngle function return an int instead of a float. Lul. Big problem when working with radians.

>>62232709
I'm not sure what your problem is.
Put your IDT in static memory, and just use an offset to get the physical address.
For example, if you load your kernel at physical address 0x00100000, and map it to 0xc0100000 in memory, then you'd just do

(u8*)idt - (u8*)0xc0000000

to obtain the physical address of your IDT.
That's what I do.

>>62233025
It's the opposite. A class has a restricted interface, and a well designed class has clear invariants so you all of the methods on it can safely make various assumptions.

>>62233150
can you explain in layman's terms? I am just a hobbyist at this point, and have no formal training

>>62233150
the only useful classes are typeclasses

File: trashkell_on_suicide_watch.png (448KB, 580x1500px) Image search: [Google]

448KB, 580x1500px

Reminder that an obsession with safety leads to uselessness.

>>62233210
Reminder that those who do not understand history are doomed to repeat it

>>62233226
reminder that no good deed goes unpunished

>>62232856
>Can someone explain to me the point of classes?
It's a big part of object-oriented programing. Basically you define a class as a template for how something is and how it behaves then you can create a bunch of instances from that class and they'll all operate as defined but independently.

A very simple example. You can write a lamp class. A lamp has one boolean attribute, isOn, which can be true or false. The lamp class will also have one "method", a function defined inside the class, called "toggle()". When you call toggle() the instance will swap the value of isOn. If it was true, it will now be false and vice versa. You can now create as many instances of this class as you'd like. Each instance will have its own isOn value which can be modified by calling the instance's toggle() method.

The point of this is that after creating the logic of the class you can make as many instances as you want and the program keeps track of them independently and behind the scenes. You just select a lamp instance, run lampinstance.toggle() and the state of that instance is modified and you didn't even have to know anything about the instance.

>If I only make small programs (say less than 300 lines), would there ever be a point in making a class?
Depends on what your objective is. If your program is centered around creating and modifying a specific type of data then classes will help with that.

>>62233177
Say you have a datatype that represents a date.

struct Date
{
    int day, month, year;
}

Someone could easily write a function that leaves a Date in an invalid state, like the 30th Februrary.

Date my_date = {30, 2, 2000};

By using access specification in a class, you can control access to the data so that it is never left in an invalid state like this.

class Date
{
    private:
        int day, month, year;
    public:
        void setDate(int day, int month, int year)
        {
            // Add some logic here to make sure that never happens
        }
}

Generally a class is easier to work with because only the interface that is useful and expected is exposed, and you can't (accidentally or intentionally) poke the wrong bits to make it break.

>>62233262
>object-oriented programing
Found your problem.

>>62233177
He's basically saying that classes are a poor man's dependent types.

>>62233244
reminder that if you aren't programming functionally, you're programming dysfunctionally

>>62233283
reminder that functional programming isn't just using maps and filters sometimes

reminder that js and erlang are not functional languages

>>62233210
Safety is only one aspect of code that is easy to reason about.

>>62233292
reminder that using higher order functions is always funtional programming even if it isn't all or even most of it

>>62233292
What is erlang if not a functional language?

Is it enough if renderer supports gles2 and gl3?

>>62233338
Enough for what?

>>62233332
object-oriented, in the original sense not the java sense

File: Bobdobbs.png (37KB, 216x324px) Image search: [Google]

37KB, 216x324px

Anyone here has worked with IoT (Internet of Things) and Fiware?

>>62233346
to support all relevant devices?

>>62233362
Depends on what you consider relevant.

>>62233177
>>62232856
At least go learn OOP before asking those questions. You either didn't even tried to understand the theory or are just trying to start a flame war around the paradigm

>>62233377
DON'T learn OOP. You'll only end up having to unlearn it later.

>>62233332
Oh shit so close

>>622333333

>>62233386
see

>>62233386
Don't listen to this retard. OOP is industry standard for a reason, but neets don't like it because they struggle to manage it properly.

gimmick idea: a C interpreter

>>62233407
>OOP is industry standard for a reason
And that reason is marketing, not merit.

>>62233417
another idea:

what has not been done in the field of programming?

i mean, everything worthy is already programmed

File: 1443571065645.jpg (12KB, 480x360px) Image search: [Google]

12KB, 480x360px

>enforcing invariants by limiting the operations available instead of just enforcing the invariants
OOP was a mistake.

>>62233262
>>62233270
ah, okay, so it sounds like its used mostly for organization in cases when I would be using the class many times, as I could make a change to the class, and it would trickle down.

>>62233422
it's industry standard because FP is too arcane for normalfags and procedural scales terribly with project size

>>62233417
There already exists multiple of those and it's not gimmick.

>>62233377
I know how to program in an OOP language, I just don't any of the theory behind it or what OOP is outside of the fact that everything is an object. Is this an issue?

What's so good about python?

>>62233450
FP is not arcane. The problem is that most universities teach OOP and not FP.

>>62233477
So much for Haskell being academic.

>>62233491
Haskell is mostly used at the graduate level.

>>62233471
Nobody can agree what OOP means anyway. Smalltalk ain't Java.

>>62233456
think i should do it in this doge autist's c parser-combinator or something else? http://www.buildyourownlisp.com/chapter8_error_handling if something else what?

>>62233491
I'm thinking mostly of undergraduate programs. There are occasional dots of ML, Haskell, and Scheme in a sea of Java, C#, and Python.

>>62233525
>http://www.buildyourownlisp.com
it's incredibly tasteless and puerile though

>>62233491
see >>62233531

universities teach OOP to brainlet undergrads

>>62233471
The theory behind OOP is that encapsulation and inheritance are a panacea.

>>62227640
Trying to learn C++ templates. What a fucking clusterfuck.

>>62233572
i bet you haven't even scratched the surface

Why is /dpt/ is so assblasted about people with actual degrees?

>>62233590
I haven't seen that on /dpt/ much

>>62233590
t. oop brainlet
your degree is worth as much as an arts degree

>>62233578
The fact that I suddenly have to implement shit in header files isn't helping C++'s case. Java generics are much better than this.

>>62233631
C++ templates and Java generics are completely different features. Java generics are more similar to void* pointers than to templates.

File: 1500325010578.png (117KB, 372x351px) Image search: [Google]

117KB, 372x351px

>>62233621
At least I have a degree

Anyone recommend a good book to understand the theories behind programming and programming languages, and their history? I want to understand what is actually happening when a write a program besides just knowing "input, ???, output". How can I learn the ??? part?

>>62233558
i wouldn't be building a lisp, i'd be using his parser combinator to do a C-like interpreter

>>62233631
see >>62233648
Also if you know what templates are going to be instantiated ahead of time you could always do it in a .cpp and use explicit instantiation.

>header files
Is this 1970?

>>62233621
>t. minimum wage McDicks worker

>>62233664
you're not actually looking for programming language theory, which is an abstract logical view of things. you're looking for compilers and assembly. assembly language step by step on the linux environment is a good book for that

>>62233648
>>62233676
They are similar features done differently. I don't give a shit how it's done, I'm not making these languages.

>>62233729
You should give a shit or else you're getting exactly the wrong idea about what your code means.

Rewriting the word/character/newline counting program from K&R by myself (it also counts tabs).
Why does K&R use two variables to determine whether or not we're "inside" or "outside" a word when that is completely unnecessary and doesn't even make the program substantially shorter?
Here's my version:

#include <stdio.h>

int main()
{
    int line = 0;
    long character = 0;
    long word = 0;
    int tab = 0;
    int input;
    while ((input = getchar()) != EOF) {
        if (input == '\n') {
            ++line;
            ++word;
        }
        else if (input == ' ') ++word;
        else if (input == '\t') {
            ++tab; 
            ++word;
        }
        ++character;
    }
    printf("Number of lines: %d\nNumber of characters: %ld\nNumber of words: %ld\nNumber of true tabs: %d\n", line, character, word, tab);
}

Here's K&R's:
https://pastebin.com/eACxCX9p

>>62233729
>They are similar features
Incorrect.

>>62233664
>I want to understand meme
Thats absolutely fine but get a book on actual programming so you can learn how to write one first. Once you have done that come back for book recomendation on 'understanding...'

>>62233768
run your program on a file exclusively containing spaces

>>62233807
Oh, right. Thanks.

>>62233775
I can already create programs. I learned all the practical stuff, but didn't learn any of the theory. Also, what's your opinion on this? What makes it a "meme"?

>>62233768
So it will say this has only one word

Hello world

And this has five words?

     Hello

File: Cbtfo.jpg (19KB, 669x55px) Image search: [Google]

19KB, 669x55px

Will C ever recover?

>>62233902
Depends on how often you use it.

>>62233902
Then you put that program into a computer and run it a billion times. Yes, it's faster.

>>62233902
There's nothing you can write in one in java that would take month to implement in C, though.

>>62233902
If I spend a day writing something in Haskell and it takes 4 seconds to run, and you spend a week writing the same thing in Java, followed by a week fixing it, and it takes 4 seconds to run, is Haskell really useless?

>>62233902
>t. poorfag who will never get rich off mining bitcoin

>>62233902
people who don't know see have some weird idea that C's a mythically arcane and difficult language, just because they can't look at svprintf and see what it does spelled out in english words for them

>>62233902
Its likely C code will be run thousands of times on millions of low powered hardware units (washing machines, fridges, toasters etc), that 0.05 second difference will add up fast

While your pajeet code will probably crash on the first run and will be discarded

>>62233974
How the fuck do you still mine bitcoins? I thought that big corporations with their super computers have taken over

>>62233902
>>62233937
>>62233942
>>62233956
>>62233972
>>62233974
>>62233990
>>62234010
amazing
>1 post and then 7 replies all in a row

>>62234012
>>62233974
isn't mining bitcoins like a lottery now instead of a steady income stream? since they're so expensive with so many people mining?

File: you.jpg (67KB, 500x354px) Image search: [Google]

67KB, 500x354px

>>62234032

Why most of supercomputers operate on linux?

>>62234112
they spent all their money on hardware and can't afford windows

>>62233902
Programmers are douchebag, they value saving a month of their time and sacrificing thousands of hours in user times with their laziness.

>>62234112
windows is bloated shit for end users

>>62234145
They value their own time. User time is someone else's time. It's like socialism: I's super easy to be generous with someone else's moneys. You're far more reluctant to give up your own time or money.

>>62234012
You can still mine as an individual, so long as you don't have to pay for electricity

>>62234034
The price has skyrocketted, making it more profitable by the day. Not a steady income by any means, but this shouldn't be your only source of income

>>62234176
how many bitcoins would you get a day on average with a standard pc?

>>62233902

Assuming a month is 31 days, the time spent running the C program will be worth the time developing it after 51.84 million runs. If this is a program that is run many times by many people, C would be more worth it. If it is only run a few times... why not use a Ruby or Python script that takes 1 second, but take only a half an hour to develop?

>>62234145
not to mention most of them write shitty programs so that they can stay employed longer, as the only person that can debug it well is the original programmer

>>62234195
>Ruby or Python
Haram

>>62234195
The development times being quoted are bullshit.

>>62232946
>Why would you need this integrated?
debugging/profiling is faster

have you ever used qtcreator?

when i run shit under calgrind from qtcreator it not only shows me a table with caller/callee and cost information, but it also shows the cost of each function near the line numbers

when i run shit under memcheck from qtcreator it'll show the list of event (errors/warnings) with clickable links to the source lines

--
for the last two weeks tho i've used asan/perf more than memcheck/calgrind

>>62234195
There are around 21 working days in a month, not 30.

>>62234188
Depending on the cyptocurrency mined, and given a high end GPU, you can earn about $1.50 a day. Pretty much shit, but if you don't have to pay electricity and only run it when at work and asleep, you can recoup the price of the graphics card in a year or so

>>62234195
just because you don't know C, doesn't mean C programmers are unicorns

>>62234233
it also doesn't mean it takes a month to write the equivalent C for a Java program that took a day to write

>>62234200
1 rule of employment: never meet or exceed your targets.

>>62234230
that's actually a lot more than i expected. that'd be very profitable if you could hijack people's computers

>>62234243
the time difference is most pronounced for very small programs. ie scripts. which is why they're called scripting languages

File: old_hag_langs.png (173KB, 600x355px) Image search: [Google]

173KB, 600x355px

>>62233902
Do Javatoddlers really have this much trouble writing C code? What a fucking joke lmfao.

Is there any guarantee that on GCC, the sizes of the pthread structures will always be the same?

I have in which the header defines some structures, some of which have as members some pthread structs (or HANDLEs on Windows).

I would like to have the header so that the user who includes the header doesn't need to also take in Windows.h or pthread.h, so I thought I could just declare the thread and mutex structures as char buffers or something like that in the header, then include pthread.h and Windows.h in the .c file instead, where I would cast the byte buffer into the correct type. But the problem is that while HANDLEs on Windows fit in 64 bits, the pthread structures are actually quite large.

Any suggestions? Just don't want to clutter up the thing from the user's perspective with the pthread/Windows header stuff or increase their compilation times needlessly.

>>62234319
We're talking about Java and C

Last I checked Java isn't a scripting language

>>62234246
Well if you create an app or something that lets you use the processing power of the phone and split the proceeds, you could possibly make a decent amount of money off mining. Even if you were to get 1000 people to download, and you make $0.4 a day, that would still be about $120 a day after all taxes. Not bad at all

>>62234372
"I have a library" was what I was going to say. Anyway...

>>62234376
ruby mentioned ruby and python, which are both scripting languages

>>62234395
My post specifically referred to the times mentioned for C and Java

File: redis-basics-24-638.jpg (56KB, 638x359px) Image search: [Google]

56KB, 638x359px

In Python, is there a data structure that provides similar functionality to the Sorted Sets of Redis? (i.e. mapping like float->object that allows you to do range-y operations: trim within a range of floats, select a range, etc.)

>>62234214

I'm willing to bet the runtimes are too. The JVM's startup time tends to show in smaller programs.

>>62234226

Then it would only be 34.56 million runs. Here I am assuming one month means one month of working days and not just one month including breaks.

>>62234233

>just because you don't know C
I made a barebones DNP3 server that can run in a freestanding environment. Piss off.

>doesn't mean C programmers are unicorns
Pretty much everyone who isn't a webdev knows C in some capacity.

>>62234372
Use a language with modules and this isn't a problem.

>>62234444 (checked)
>I'm willing to bet the runtimes are too. The JVM's startup time tends to show in smaller programs.
That's why Java9 has AOT.

>>62234473
It does?

>>62234416
don't care

>>62234444
this isn't the webdev general. are you lost?

Is this guy right?

https://www.youtube.com/watch?v=QM1iUe6IofM

>>62234517
stop posting your video

>>62234498
http://openjdk.java.net/jeps/295

>>62234517
>serious efficiency problems with functional programming
Dropped.

>>62234517
I'm not going to watch a 45 minute long video of an autist talking about OOP. Give me a TL;DR.

>>62234516
He's an utterly moronic tripfag. What do you think?

>>62234627
tl;dr fine grained encapsulation (i.e. objects, not modules) doesn't work.

>>62234554
It's not mine, I want to be educated and I'm not sure if I should listen.

>>62234627
Just start at 18:09 and watch for 5 minutes. Make sure your adblock is on if you don't want him to profit.

>>62234379
You need full high end gpu farm to even get few dollars a day from bitcoin mining. Even if you get 1000 people to download your app, running mining code will drain the battery in an hour.

If you can hack 10000 gaymurs, create a botnet then it might be worth it other wise its not

Another way is to invest few million into setup and then hope the price doesnt crash

>>62234667
>>62234683
The video is a waste of time and effort. He's retarded. Good OOP code is just as good as good FP code. Some problems benefit from paradigm X while others benefit from paradigm Y.

File: toolbelt anon.jpg (112KB, 1280x720px) Image search: [Google]

112KB, 1280x720px

>>62234827
[citation needed]

>>62234685
What? I get $2.50 a day from a single (1) 1080.
My phone could make 0.01 per day, roughly. If he took half that's half a cent per day per user. Not too bad, especially when paired with ads.

>>62234850
OOP languages top the TIOBE index
OOP is what makes the world go round

>>62234889
>OOP is good because it's popular

>>62234889
Which OOP languages would those be?
I hope you don't mean Java, C#, or C++, which are for the most part structured and procedural, and add tiny bits on top of C?

>>62234895
>I'm a special snowflake
>>62234906
Java, C# and C++ are all object oriented.

>>62234947
Being skeptical of popular->good is not the same as saying popular->bad or unpopular->good.

>>62234961
So why is OOP bad again?

>>62234961
Did you just try to seem smart?

>>62234961
Hating OOP because it is popular is just as bad.

>>62234985
he's right though, about popularity and value
he's wrong about OOP being bad though

Is there a way to check all contents of an array in C on demand?

>>62234947
>Java, C# and C++ are all object oriented.

Keep telling yourself that faggot
>i-it has x.f() method syntax!!
>i-it has virtuals
>the other 99% of non-oop features don't matter!!

>>62235009
>on demand?
what's that mean

>>62235016
Go fuck yourself. They're not as `pure` as SmallTalk, but they're still OO.
You're like those sad kids that claim Haskell is the only FP language out there.

>>62234978
Because it provides a single tool (objects) that propose to handle both ownership and abstraction while being ill suited to both. Look at how pervasive shared mutable state and design patterns are in production OOP code.

>>62234985
No, I'm being logical in the face of a logical fallacy spouting retard.

>>62234996
Who is doing this?

Is it possible to have an arm co-processor on a pcie card or something so I can do some hardware ARM dev or do i need a raspberry pi or whatever

>>62235040
>use a language that is only 10% OOP as a standard of OOP and accuse me of purity spiralling
holy shit get a grip

>>62235039
At any time.
To check the elements

>>62235073
Show me the ISO standard of OOP.

>>62235104
use a loop?

>>62233460
Saying OOP sucks is like saying capitalism sucks. You're right but you sound like a soviet for saying it.

>>62235042
dpt hates a lot of things that they don't understand simply because it is popular to hate it on dpt

>>62235160
Hating something because it is popular to hate is different than hating something because it is popular.

>>62235117
Show me the ISO standard of FP
Oh I guess they're all FP languages then!

You are so desperate it's hilarious.

OOP features of C++:
>virtuals (aka HOF + function member, alt. typeclasses/concepts/traits/interfaces which are a million times more flexible and useful)
>methods (aka just x.f() syntax)
>access modifiers (aka what non-OOP achieves with modules)

OOP features of C#:
>virtuals (aka HOF + function member, alt. typeclasses/concepts/traits/interfaces which are a million times more flexible and useful)
>methods (aka just x.f() syntax)
>access modifiers (aka what non-OOP achieves with modules)

OOP features of Java:
>virtuals (aka HOF + function member, alt. typeclasses/concepts/traits/interfaces which are a million times more flexible and useful)
>methods (aka just x.f() syntax)
>access modifiers (aka what non-OOP achieves with modules)

meanwhile virtually everything else is non-OOP

>>62235154
i hate analogies and people who use them
always seem to think themselves SO clever, meanwhile analogies prove nothing

analogies are stupid

>>62235009
Loop through all elements.

>>62235181
>>62235122
like wtf why can't I just hoover my mouse over it to shot me the elements? this is dark souls of programming

>>62235178
Using an analogy to prove a point is like trying to explain an abstract idea using real life and relatable examples.

File: scr.png (28KB, 808x627px) Image search: [Google]

28KB, 808x627px

I did a maze generator using a recursive procedure

File: 1412164398657.jpg (95KB, 480x480px) Image search: [Google]

95KB, 480x480px

>>62235240

>>62235240
>Using an analogy to prove
no

>>62235218
In what context?

>>62235240
>like trying to explain an abstract idea using real life and relatable examples
That approach doesn't work.

File: spoonfeedmepls.jpg (10KB, 194x249px) Image search: [Google]

10KB, 194x249px

>>62235218
>like wtf why can't I just hoover my mouse over it to shot me the elements?
what did he mean by this?

>>62235249
looks nice anon

>>62234372

pthreads might be different depending on platform. They run on Windows too, so don't bother with HANDLEs, just use winpthreads. #including a single header doesn't take that long, especially if it's just regular C and not C++ with templates in it.

>>62234516

I don't do webshit, I do security research.

>>62235174
Normally one would say a language is OO if it has objects which contain attributes and methods.
Obviously a turbo autist like you would claim that only SmallTalk is the one and true OO language.

>>62235278
The overwhelming majority of those three languages are NOT oop.
To say that adding some tiny bit on top of that language makes it an OOP language is ABSURD.
I could just as easily say that now they've all got lambdas (an unquestionably functional feature), they're all functional languages.
Wouldn't be true.

>>62235275
Which is better, RSA or DSA?

>>62235296
by "overwhelming majority of those three languages" I mean of each of them, not 2/3

>>62235305

My main focus is on intrusion detection, not encryption. I've studied cryptography before, but we did not really cover DSA. Looking into things, we've had larger RSA key lengths broken compared to DSA, but that might just be because RSA is in use more.

>>62232369

/* valid */
int values[9] = {0, 20, 20, 30, 80, 70, 0, 0, 20};

/* not valid */
int values[9];
values = {0, 20, 20, 30, 80, 70, 0, 0, 20};

    int values[9];
    if (something) {
        int xval[9] = {0, 20, 20, 30, 80, 70, 0, 0, 20};
        for (int i = 0; i < 9; ++i)
            values[i] = xval[i];
    }
    else {
        int xval[9] = {0, 40, 30, 50, 20, 20, 0, 0, 80};
        for (int i = 0; i < 9; ++i)
            values[i] = xval[9];
    }

>>62235305
Use elliptic curves my man

>>62235483
oh i see
i just wrote it out values[i] = number and so on for each case
seems about as elaborate
either way, it's a hack, but it's ok

Starting tomorrow, I'm going to be working on porting an app to iOS. Have to do adjustments to refine it for an iOS experience and implementing ios stuff to it. Probably going to take a couple of weeks.

What's a good way to prepare for this mentally? Clean the room? Arrange stuff so I can take off full force starting in the morning? Grab some donuts?

File: cut_churro.png (47KB, 436x297px) Image search: [Google]

47KB, 436x297px

Whats the easiest way to do hardware ARM-Cortex programming from a windows machine?

Trying to avoid that Debian/RaspberryPi/etc type of stuff and do some low level stuff

>>62233108
The problem wasn't finding the IDT. It was figuring it had to be loaded with the virtual address (harder when every faggot calls it "linear" except we're in long mode so this doesn't apply anymore). Every reference tells it's the same as the GDT but the GDT is first loaded with a virtual address equal to its physical since you need to load it to even start going into long mode. Good thing I fixed it all. I just nuked all my data segments with the null descriptor and it fixed everything. God bless long mode.

>>62235542
Get a switch dev kit. lulz

New thread:

>>62235587
>>62235587
>>62235587

/* nigger faggot */

>>62233902
>Java
you mean Go

>>62228679
>clojure
Not even in the top 20 over at tiobe