/dpt/ - Daily Programming Thread

File: .lain..jpg (54KB, 546x896px) Image search: [Google]

54KB, 546x896px

>>62288936

Haskell!

What's the best programming language, and why is it a functional one?

>>62289242
see the post above yours

>>62289242
I saw an HTTP server written in prolog once, I don't understand you people

When books say that function variables are declared "on the stack", do they mean they're stored in the computer's actual hardware stack, or that they're just represented in a stack data structure?

I'm just wondering, because if you try and access the first variable declared in a function, does the OS (or whatever deals with the low-level shit) have to pop all the other variables off and then push them back on?

>>62289313
every program running has it's own stack, there's a bit of boilerplate ASM that gets created to make this happen even if your C program does nothing but return 0.

File: 1502119898289.gif (1MB, 500x225px) Image search: [Google]

1MB, 500x225px

Back in the day I used to draw in BASIC with commands like POINT, LINE and such.
Does there exist a Python library which would let ne do this? I'd like to save my pictures in some vector format, though.

File: p0U4veG.jpg (229KB, 1366x768px) Image search: [Google]

229KB, 1366x768px

I need some help senpais i'm learning c and the book gives a program for listing prime numbers, I wrote my version and the console output was wrong after changing things and making sure it was perfect console output was still wrong. So i copied the code from the ebook still dodgy console output what's wrong?? Here is the code

// Program to generate a table of prime number

#include<stdio.h>

int main (void)
{
int p, d;
_Bool isPrime;

for ( p = 2; p <= 50; ++p ) {
isPrime = 1;

for ( d = 2; d < p; ++d )
if ( p % d == 0 )
isPrime = 0;

if ( isPrime != 0 )
printf ("%i ", p);
}


printf ("\n");
return 0;
}

Bottom left of image is console output

>>62289470
lacking some braces there

File: 1502076029715.gif (929KB, 500x388px) Image search: [Google]

929KB, 500x388px

Do people need to learn how to program?

File: 1504711784563.jpg (274KB, 750x750px) Image search: [Google]

274KB, 750x750px

>>62289242
None, they're all retarded and chaotic. Haskell being """pure""" is just a meme.

>>62289501
can you check my original code because i'm pretty sure that is right but had same console output

// Program to generate a table of prime number

#include<stdio.h>

int main (void){
_Bool isPrime;

for ( int p = 2; p <= 50; p++){
isPrime = 0;
for ( int d = 2; d < p; d++){
if (p % d == 0){
isPrime = 1;
}
if (isPrime != 1){
printf("%i ",p);
}
}
}
printf("\n");
return 0;
}

>>62289516
My math bud says that "programming is basically accounting, everyone can do it after a week or so", I don't consider that bullying because he's extremely autistic

>>62289529
Haskell is statically typed functional language. There is nothing special about that, aside from utility. Dont know why you're talking about "purity," whatever that means.

>>62289565
what is accounting even?
it's just counting and making sure money wasn't lost between the cracks

You can fill the array manually while declaring it, in C, like this

 int array[] = {4,6,7,3};

can I fill it in the same manner but form a file without using loops?

>>62289585
yes, but why would you want to?

>>62289583
People rather use Excel than programming to perform accounting though.

>>62289558
You're resetting the isPrime state, notice line 14.
https://ideone.com/kZOiqP

>>62289583
>what is programming even?
>it's just counting and making sure data wasn't lost between the cracks

File: 1461871914470.gif (893KB, 400x554px) Image search: [Google]

893KB, 400x554px

>>62288936
Trying to threadsafe a mutex method with a c# control. Failing miserably.

>>62289585
No, you can only initialize arrays with constant literals. I/O is not constant by definition, so you still have to parse the stream before filling your variables.

File: 12894883_1200x1000_192.jpg (102KB, 1200x800px) Image search: [Google]

102KB, 1200x800px

https://www.youtube.com/watch?v=JpkjOkiu-fA

ironic that an animu poster is such a SJW

>>62289664
In this case, programming in C can't ever be called programming.

>>62289581
purely functional i would guess
WHICH IT IS

>>62289719
haskell isnt pure.

>>62289726
What do you mean by "pure?" Haskell is just a statically typed functional language.

>>62289726
whats your pleb definition of 'pure' in the context of (functional) programming languages?

>>62289726
It's pure if unsafePerformIO is used correctly or not at all. Otherwise laziness would not work.

>>62289744
no mutation of external state, deterministic execution, etc.

CPUs don't work this way, so claiming to do pure programming is fucking stupid.

>>62289699
check this one
https://www.youtube.com/watch?v=3hviD9fuuSc

>>62289761
wrong.gif you can model proper hask code as a purely rewriting system, no von neumann bs

>>62289761
That is not what purity means. Purity means you get equal output given equal input.

>>62289585
from a file? if you really do mean "in the same manner" (as in direct initialization) then... sort of, haha. technically you can add a file with just comma-delimited values in it to your include directory, and then do

int array[] =
{
#include "values.txt"
};

the values will be included at compile time, though, and changes to the values will require rebuilding, so this probably isn't a good idea

>>62289774
the only video from the conference that i couldn't finish, only because the accent was so horrendous

>>62289470
Why not VSCode?

>>62289908
What is vscode?? I've wanted to learn c for a while want to get into 3ds and vita homebrew

>started CS
>never did programming before
>just read the book and do all exercises before labs
>have almost no problems during lab because ironed out syntax errors in book exercises
>people who sit near me struggling to write simple shit and say stuff like
>"wow, anon, you must have a lot of programming experience"
>hear other people say
>"wow bro this is so hard I don't understand anything lol"
Why do these people even attend collage?

What do I use to git gud at C++?

>>62289989
Learn Rust instead.

>>62289977
parents forced them, computer jobs are all the rage
not everyone is cut out for college
the sad part is everyone will automatically assume you're as stupid as the lowest denominator of compsci student

>>62290009
But I want to learn how to handle memory allocation.

>>62289908
>Why not VSCode?
shills please leave

>>62290012
Makes me so mad when I see CS students with computers

You don't use a computer to learn CS

You use pen and paper

CS is proofs, not programming

You're not proving anything hard/complex enough to require any automation in an undergrad program

>>62290118
maybe in 1960

CS is a programmer degree now.
You might have math up to physics, but they still expect you to learn how to program, because otherwise you're fucking useless.

>>62289989
>>62290071
cppreference.com
Herb Sutter's GotW series
Jason Turner's C++ Weekly series
CppCon and BoostCon presentation videos
Books by Stroustrup, Sutter, Meyers, Alexandrescu
to name a few

>>62290118
I don't understand what you're saying.
Everyone uses a computer in collage. Writing reports, mail. And the how are you supposed to program without a computer? You're not making any sense.

If you're talking about math, then yes there is 0 need for a computer.

Trying to figure out can bus with stm32 and its fucking weird

>>62290198
CS is a subset of math

>finally figure out how to work with file data embedded into a program

I feel like a fucking god

>>62290222
CS is the intersection of math and programming.

>>62290223
rust has a macro for that

>>62290118
>>62290222
>muh dreamland autistic purity
>muh made-up completely subjective redefinition of common term
Classic hardcore autist behaviour. Ignore it.

>>62290233
On the contrary, it's the intersection of art, Engineering, and math. Pretentious fuck

>>62290071

I would recommend getting familiar with C before you start writing C++.

C++ started as a fancy macro set on top of C. Once you understand C structs and memory use, then C++ is both easy and clearly seen as completely fucked up from an ecosystem standpoint.

Then you'll probably want to learn Rust. If you're wanting to learn C++ for game dev, you'll still be very well served by learning C first.

>>62290244
that's cute. personally i prefer to generate my data procedurally at compile time with constexpr

>>62290292
>then C++ is both easy
hahahaha
>and clearly seen as completely fucked up from an ecosystem standpoint
sure, when you don't actually know it

>>62290244
Why do people still talk about Rust?
Is it some sort of thing where people think it's better because it goes full babymode?

This is a very basic snippet from the python script I am writing:

running = True
while running
usri = input("Do you want to proceed? (y/n): )
if usri == "y":
function()

It says the namerror: y is not defined. I looked it up on my favorite search engine and learned that this has to do with the scope of the variable. But, usri = input and the if-block resides in the same block, how is this possible?

>>62290282
>>62290233
>>62290267
>>62290222
>>62290118
>>62290146
>>62290198

> semantics

Cringe so hard. Who cares? Everything is defined differently by the *context* in which you're viewing it.

CS is whatever the university system decides it is. The science of computation. Nothing more, nothing less.

>>62289470
jesus fucking christ if that's your top just kill yourself immedietely

>>62290292
>C meory use
What you people mean by this? Declaring variables, manipulating arrays etc right?

File: PhilFish-610.jpg (140KB, 610x345px) Image search: [Google]

140KB, 610x345px

>>62288936
>Avoided Java for the longest time because people shit all over it
>Using it for my data class
>Loving using it

Why'd you guys lie to me?

>>62290364
Semantics are important in CS. You should know that.

>>62290383
allocation, stack and heap. pointers.

>>62290091
>FOSS
>shill

????

>>62290418

Sure. *In* CS. Not when discussing CS itself.

>>62290198
>If you're talking about math, then yes there is 0 need for a computer.
lot of my math professors use CAS (computer algebra systems) and matlab/octave for numerical number crunching.

computational maths is getting bigger and bigger.
you can use it for example to sanity check some idea you came up with. or look at large numbers of examples to try to find patterns

tl;dr "no"

why's it always print yes?

read input
if ((input == 'Y' || input == 'y'))
then
        echo "YES"
else
        echo "NO"
fi

File: Azusa 415.png (910KB, 835x835px) Image search: [Google]

910KB, 835x835px

tfw banned from using rtti ever again on production

>>62290495
language is bash btw

>>62290412
it's unnecessarily verbose and has "corporate" written all over it.

what other languages do you know of that you can compare it to?

>>62290363
Count the quotes.

>>62290515
this is my first language

>>62290412
Java is a creepy fucker because it looks like a great language until you're sitting on a few thousand lines of it. Then you realize you're in a curry smelling kafkaesque code monkey distopia.

>>62290507
good

>>62290315

Maybe you misunderstood.

C++ is easy to understand.

You just have to untangle 30 years of ecosystem changes to be productive in it. Multiple standards, multiple template libraries, build systems that are absolute anus (looking at you, autoconf; or hell, CMake which has the shittiest documentation I've ever seen), etc.

Fundamentally simple. Just shit on top of shit.

>>62290412
Java is the best programing language, hands down.

>>62290495
>>62290508

This will always echo YES because you can put anything inside those parens and it'll evaluate truthily.

>>62290520
I mistyped it here. It is correct in the original script.

>>62290412
C# is better tho. wouldn't make much difference earlier on, admittedly, but check it out down the line sometime if you ever feel like you're hitting a ceiling

>>62290495
http://tldp.org/LDP/abs/html/dblparens.html

>>62290495
use square brackets for these conditionals, see also

$ man test

I think double braces are used for bash-arithmetic, but maybe I'm wrong. Also string comparison is done via "=", not "==" (at least in test)

>>62290582
the C# ecosystem is full of proprietary tools, the Java ecosystem is full of open source tools. And dont get me started on trying to use mono if you're on linux.

My CompSci starts with C and then Java right after and no C++ at all.
I don't get it, wouldn't C++ be better since we already learned C?

>>62290615
>>62290540
Not that guy, but basically that. Your teachers are lazy and C++ is a two book subject. Like physics. Good idea to learn for yourself though.

>>62289800
Then all languages are pure. Given the same CPU/memory/HDD state they will yield the same value

>>62290640
A program might be pure but purity of individual functions is much more rare (except in Haskell, for instance).

>>62290615
>>62290637

I am that guy and I'll echo the sentiment. C++ is too complicated to teach students. There's so much history to unfold that it's simply not possible to teach it *well* to a class of undergrads.

I honestly think that this is why it's probably on its way out. Ain't nobody got time for that.

>>62290598
>>62290588
>>62290577
cool thanks

>>62290649

How does GHC work?

>>62289313
On x86, , "the stack" is just RAM like any other.

It's not magic in any way. It's exactly as if you had implemented a stack using an array and a size. Pushing is array[++size]=42; Popping is --size;

And of course, reading the Nth value from the top is array[size-N], no popping required.

>>62290578
>not copy pasting
First year is going to be hard. Count colons.

saying my program is gpl but I'm not going to release the source code when anyone asks

>>62290707
based

>>62289398
https://stackoverflow.com/questions/4071633/python-turtle-module-saving-an-image

>>62290640
Not at all. Take getch().

>>62290682
How doesn't it?

>>62290743

Ha ha touché.

>>62290757
I mean, equal inputs leading to equal outputs is not incompatible with anything. You just have to model state and effects as input/output.

>>62290702
https://gist.github.com/anonymous/afc4f8352f4592fd4ecf980fd4278be1

Here is the code. I don't see anything wrong with colons. Please elaborate.

>>62290707
i release all my software as free software but the precompiled binaries have botnet inside them

>>62290682
It rewrites your shit into Core, then into C, then it compiles to asm.

i fucking hate bash

>>62290788
It's basically a pure function from strings to assembly code.

>>62290775

if usri == 'y': #continue if yes

jej

>>62290803
me too

why is it telling me command [[y not found

read input
if [[$input = 'y']]
then
        echo "YES"
else
        echo "NO"
fi

can't find y lads

>>62290894
Stop using shit language

>>62290775
Missing colon in your original post you fucking moron. If you need help on the first weekly assignment better drop out.

>>62290882
Please paste this shit into shellcheck before asking

>>62290882

You need to quote $input.

It's interpreting it literally and then executing it.

>>62290932
Except I am not taking a course. I'm trying to follow a book I found at Python's website.

>>62290945
I understand. With that intellect I would also save the college money.

>>62290882
try spacing things out a bit

if [[ $input -eq 'y' ]]

>>62289565
probably because he cant do either one of those

>>62290937
never heard of it
>>62290938
this wasn't it
>>62290987
this was it

>>62290986
It's my first week with programming you stupid nigger.

>>62291002
>never heard of it

You're missing out, bruh. You could be getting warnings like this straight in your local editor:

Line 2:
if [[$input = 'y']]
   ^-- SC1035: You need a space after the [[ and before the ]].

I want to program an algorithm to resize an image, however I want to see what's the most efficient way to do so. I have a way right now (pseudocode) but I want something more efficient:

 int height = image.height; 

 int width = image.width;


  if ((height>=1000) || (width >=1000))  
  {

     image.height =    image.resizeHeight(image.height * 0.25);
     image.width = image.resizeWidth(image.width * 0.25); 
  }

//this could work for some image sizes, but I would have to iterate for images with a larger size

//in any case what would be the best max size of an image? 2000 x 2000?

Your thoughts /dpt/?

>>62291004
Then I would suggest you give up now. You will never succeed Tyrone. If you can't resolve a simple issue, does not know how to copy paste, and then seek help on an anime picture forum? Just my 5 cents.

>>62290540
>C++ is easy to understand
nah, man. you'd only say that if you were unaware of a lot of the details

> Multiple standards, multiple template libraries, build systems
the standards revisions aren't that hard to stay on top of in themselves. it's the fact that they (along with the foundational language) culminate into a metric shitload of interconnected details even if you nix all the shit that's been deprecated or effectively made redundant. and build systems are kind of another matter altogether, but CMake isn't even that bad. if that's the worst documentation you've ever seen you are one lucky son of a bitch

>Fundamentally simple
hell no. C is fundamentally simple. the C++ standard and Stroustrup's book are both a good chunk longer than the bible. and unlike similarly sized shit like the OpenGL reference, the standard defines a bunch of things that form often complicated relationships which culminate into weird edge cases and techniques that don't get discovered for years after they're made possible by the feature set. void_t and the detection idiom got discovered like two years ago and they've been possible since C++11. it's not the ecosystem that makes C++ an undertaking, it's the language

still sure as shit worth knowing if you wanna work in a couple of sectors, though. the few places it's still got a stranglehold, it's for good reason. its grip on those markets still looks like it's gonna hold for the foreseeable future. when it does eventually get overtaken for shit like games, i really doubt it's gonna be by any language we've seen yet today

wtf is 'green coding' and how do i get involved?

>>62291074
If an arrogant nigger like you can't point out what is wrong with a first-week-assignment tier script, then I suppose I should give up on visiting this shithole and continue reading the book.

>>62289470
this will give you the same output as the example

#include<stdio.h>

int main (void)
{
int p, d;
_Bool isPrime;

for ( p = 2; p <= 50; ++p )
{
isPrime = 1;

for ( d = 2; d < p; ++d )
{
if ( p % d == 0 )
isPrime = 0;
}
if ( isPrime != 0 )
printf ("%i ", p);
}

printf ("\n");
return 0;
}

>>62291099
I wont give you the answer straight Tyrone,you have to think for yourself.
The interpreter literally already gave you the solution. Little clue, do you have an variable named y?

>>62291180
I assign the input string (either y or n) to usri variable. There should be no y variable, what do you mean?

>>62291210
I figured it out, thank fucking god. Geany was running my code inside the python2, the script works as intended with python3 via terminal.

Best C# book for someone who already has a clue?

>>62291308
C# in Depth by Skeet
"C# 7.0 in a Nutshell" when it releases, judging by prior content.

File: 1485652958941.jpg (131KB, 591x800px) Image search: [Google]

131KB, 591x800px

Have you loved Lain today, anon?

>>62291346
I appreciate it

// the types of f and n could actually be inferred
// the type of xs is required to call its methods
map(f: F, n: UPtr, xs: *Array<A, n>) where F: Fn(*A) {
    // indices returns a lazy producer for the sequence 0, 1, ..., n - 1
    for i in xs->indices() {
        // types can define methods on their respective pointer types
        f(xs.element(i));
    }
}

>>62291078
Fair enough. I hadn't considered the metaprogramming-on-top-of-metaprogramming stuff. Maybe you're right; the language itself is complicated.

File: 1504650540582.jpg (151KB, 975x962px) Image search: [Google]

151KB, 975x962px

>need to gather statistics on a rather large data set
>decide to dump the whole thing into a postgres database on my desktop pc
>one of the tables has over 400M rows
>running a basic select query takes about 15 minutes

what the FUCK can I do to make this process less painful?
I already have an index on the primary key, gave postgres 80% of system memory, and beefed up caching.
what else can I do to speed things up?

>>62291685
buy supercomputer

>>62291685
SSD

>>62291685
Build a Hadoop cluster and run a MapReduce job over the data.

>>62291685
dumb frogposter

>>62291693
pls
>>62291705
i'll consider it if all else fails
>>62291711
>hadoop cluster
Would it have any significant performance gain over postgresql if I run it on a single machine? I don't have access to enough machines for a real cluster

>>62291685
>15 minutes
who cares, that's plenty fast
>>62291738
>Would it have any significant performance gain over postgresql if I run it on a single machine?
no

>>62291738
> Would it have any significant performance gain over postgresql if I run it on a single machine?
>>62291769
> no
just what this anon said

it's about as fast as you're going to get without splitting the work up somehow

How does CORS work. I don't even.

Goddammit

>>62291769
>>62291786
Alright, thanks for the tips.

I'll just queue up a bunch of queries and leave them running overnight

>>62291835
>browser makes XHR
>server responds
If the response's headers show that CORS is enabled for that request, the browser will pass it along to the javascript code.
If not, the browser will throw away the response and pass a 401 error to the javascript.

>>62291260
Figures you're a Mac user

alright /dpt/ let me pose a question

You have some data that you want to access as fast as possible, preferably in O(1) time. Typically, you'd want a hash map for this, but the issue is there are multiple keys, and you should be able to query the data structure with a few combinations of those keys.

A coworker has this problem, and he's just duplicating the data and using separate maps for each query type but this feels lazy. My initial idea was having a map for each combination of keys, and have the value be a pointer to the actual data, but this still feels sloppy.

Any thoughts?

>>62291936

Time memory trade off. If constant time is what you want, duplication is what you're going to get.

>>62291936
>>62291947

Also yes, that solution is fine.

I have a Rust question.

According to their book, Vectors and HashMaps have to be homogenous with regard to type. You can use an enum to store arbitrary types, and this tells the compiler what it's going to allocate (I'm assuming this is similar to a union in C).

Does this mean that if I have one huge element in my enum and one small one, a large number of small elements will take up a huge (theoretical) amount of memory?

>>62292030
s/element/type/g

sorry

>>62292030
That would make sense but I'm not sure myself

File: 1464955242257.png (911KB, 782x935px) Image search: [Google]

911KB, 782x935px

>3D arrays***&

>>62291891
Debian GNU/Linux operating system with LXDE, acktually.

>>62292209
stuff[x + MAX_WIDTH * (y + MAX_DEPTH * z)]

>>62292030
Yes, that's correct. If you want to be able to store either a large or a small element, and the large element is much less frequent, you could Box it so that its just a pointer.

enum Thing {
  Big(Box<BigThing>),
  Small(i32),
}

This should take up <=16 bytes even if BigThing is very big (big as in lots of fields, not big as in something like a vector of lots of elements)

Can anyone familiar with log4j explain to me how this is letting through error messages? The info level shouldn't let them through. And If I change it to level = error (which should let both though) it only lets error messages through.

<?xml version="1.0" encoding="UTF-8"?>
<Configuration>
     <Appenders>
         <File name="A1" fileName="C:\opt\logs\Bitacora.log" append="true">
             <PatternLayout pattern="%t %-5p %c{2} - %m%n"/>
         </File>
         <Console name="STDOUT" target="SYSTEM_OUT">
             <PatternLayout pattern="%d %-5p [%t] %C{2} (%F:%L) - %m%n"/>
         </Console>
    </Appenders>
    <Loggers>
         <Logger name="org.apache.log4j.xml" level="info">
             <AppenderRef ref="A1"/>
         </Logger>
         <Root level="info">
             <AppenderRef ref="STDOUT"/>
             <AppenderRef ref="A1"></AppenderRef>
         </Root>
    </Loggers>
</Configuration>

File: 1467167537894.jpg (2MB, 2000x2000px) Image search: [Google]

2MB, 2000x2000px

>>62292234
>allocating 3D arrays***& in the heap

>>62291947
Yeah but having the memory duplicated for each way we want to access the data seems like wasting memory.

>>62292249
info = info and above. not sure how to exclude error but let info pass though

Forgive me if the question is naïve, but is there a language where you can metaprogram inline using the same or similar syntax? For example, surrounding all metaprogramming with square brackets. Inside, "if" acts like "static if", "assert" acts like "static assert", etc.

>>62292243
Thanks anon

Not sure yet where the compiler is smart and where the compiler is *really* smart

>>62292250
That was for indexing, you inbred. Put it on the stack with DEFINEd MAXs if you want.

>>62292209
dumb emojiposter

File: Untitled.png (16KB, 1052x391px) Image search: [Google]

16KB, 1052x391px

>>62292266
>not sure how to exclude error but let info pass though

According to the manual the INFO level should do just that. Why is it lying to me.

>>62292272
You may like lisp.

>>62292295
you're reading the chart sideways.

>>62292316
Oh wow, I am. Thanks

File: ascended.jpg (649KB, 1920x1080px) Image search: [Google]

649KB, 1920x1080px

redpill me on reader macros

>>62292285
Generally, modern compilers don't make any optimization on data layout. Don't assume the compiler is "sufficiently smart". The smartest thing the Rust compiler does is type/lifetime inference, then LLVM does a bunch of optimizations on the instructions that are equivalent but faster (e.g. assigning stuff to registers, inlining, constant folding, and much more)

>>62292458
Unnecessary in most cases

>>62292329
no prob. log4j also understands json and yaml config formats. i find those are easier to read than xml, but that's just me

>>62292506
which cases are they helpful in?

>>62292526
Reader macros are nice because they allow the "standard language" to be defined in terms of a "core language". For instance, the main Racket language (racket/base) comes with a few reader macros:
>immutable vectors: #(x y z)
>byte strings (bytes) #"abc"
>immutable hashes: #hash((k . v) ...)
>regular expressions: #rx"[a-z]+?"
These are all implemented somewhere in the standard library as reader macros, rather than hardcoded by the compiler. Well designed languages are extensible rather than rigid.

Also, these cover the majority of cases where you want a reader macro to extend s-expressions (additional datum forms for common immutable data structures). If you create another ubiquitous data structure, I suppose you could add a reader macro. But for most cases, reader macros will be difficult to implement and impair the readable of any code that uses them, so you probably should avoid them in favor of regular macros. S-expression based macros have lots of tools (srclocs in syntax objects, syntax/parse, etc.) and can be plenty expressive.

>>62292626
That makes a lot of sense, thank you. I'm currently learning common lisp and just got to the advanced metaprogramming features

>>62292662
Racket makes it slightly harder to write reader macros than Common Lisp, and sometimes I hear Common Lisp people using that as a point against Racket, but to be honest it doesn't really matter because defining your own reader macro is extremely rare (I don't think I've ever done it seriously, outside of just testing to see what reader macros are)

Why is Java used so much in industry?

>>62293132
Because every Java developer is the same. You can replace them as often as needed without worrying. Perfect for big business.

>>62293132
because muh android

Where can I find a good tutorial for x86 assembly lads? Is assembly really that much faster than Python?

>>62293201
Says every single CIO who doesn't know a damned thing about software development and sits on their ass collecting a bigger paycheck by outsourcing to Tata or Infosys while their IT burns to the ground. I can't wait until the myth of the language skillset dies. Languages don't mean shit.

>>62293224
this all began long before android

>>62293132
Name a language and I can explain why Java is used over that language.

>Python
Java is statically typed
>C/C++
Java is garbage collected.
>C#
Java treats linux as a first class citizen.
>Haskell
Java is mature.

>>62293275
>t. micropenis javafag

>>62293244
>Languages don't mean shit
the brainlet inside is revealed

>>62293275
> Java is statically typed
Go
> Java is garbage collected.
Go
> Java treats linux as a first class citizen.
Go
> Java is mature.
Go

>>62293356
Java has generics.

>>62293353
Polyglot or gtfo. Come at me bro.

>>62293356
lol, with type erasure

excuse me while i masturbate

>>62293368
Not mutually exclusive. Knowing many languages (well) does not mean those languages are not distinguished / "mean shit". You're out of your element.
Back to work Pajeet, you don't want to be replaced now.

>>62290534
retard

>>62290507
Nothing wrong with RTTI. std::variant relies on it.
Are you also banned from using std::variant?

File: c++.png (118KB, 407x354px) Image search: [Google]

118KB, 407x354px

>C++ doesn't have ranges yet

>>62290507
what did you do?

>>62293362
btfo

>>62293275
>Java is garbage collected.
How is this an advantage?

>>62293397
> Because every Java developer is the same. You can replace them as often as needed without worrying. Perfect for big business.
This is bullshit of the highest order. *Not* every Java dev is the same and just because you're using a "distinguished" language that doesn't say *anything* about your chances of success. Do you know how many times I've seen some Oracle solution hack job put together by Indian consultants where they clearly don't understand the first thing about the tech they're programming? I'm a consultant. I see it *all the time.*

YOU are out of your element. Choosing Java nets you very little in the way of project success. I'd rather have an experienced dev team write a wrapper in the team's favorite language than have some no-talent ass clowns try to use MQ or Hibernate for something retarded for the billionth time. Fuck off.

>>62293509
Enjoy using references to pointers to pointers to pointers, faggot

>>62293519

Is that what your code looks like?

No wonder you need a GC.

Enjoy sucking cocks.

>>62293535
Nothing wrong with sucking cocks

>>62293516
>just because you're using a "distinguished" language that doesn't say *anything* about your chances of success
By distinguished, I meant something unique and beneficial for a language (e.g. Lisp sexpr's simply don't exist in Java, no matter how good of a Java programmer you think you are).

>Choosing Java nets you very little in the way of project success
I agree. It's a piece of shit corporate language, and I was never endorsing it. But it does have incredibly strong industry momentum for the reasons I specified (cheap, replaceable workers en masse as risk reduction. Throw enough monkeys with typewriters at the problem..)

Im taking an OS class.

We can do homework in java and C++/C

Am I fucked? I only know java and i dont think java can do system calls like fork()

>>62293600
>takes an OS class
>doesn't know C
You are dumb, anon.

>>62293616
WELL IF YOU NEED C TO PASS THE CLASS WHY ISNT IT A PREREQUISITE

Hi guys, I know this will be a dumb question, but i'm brand new to programming, so bear with me...

in Visual Studio 17,

A.how the hell do I rename a project? When I go to the solution explorer on the far right and right click on the project, and rename it, the files in file explorer stay the same name (where as when I rename the solution the solution name changes right away in file explorer).

B. How do I use the "start without debugging" feature on anything other than the first project in the solution? I.e. I can create 2 projects with separate .exe's but when I hit ctrl-F5 w/o fail VS only will run the first project. And if I clean the solution and rebuild only the second project, and then hit ctrl-f5, VS spits out an error and asks where the first project is and refuses to run.

>>62293688
A. When in doubt, realise that all solution/project files are in XML and text editors exist.
B. Right click project -> Set Active

>>62293600
I've never taken an OS class, so take this with a grain of salt, but I'm sure you can write a toy "operating system" in Java for such a course. Not having system calls like fork() is not a problem - those are system calls that require an OS to be running in the background! For multithreading, Java of course has it's own structures and functions in the standard library.

Is there some sort of field of thought that defines math-like concepts for functions? Setting down certain rules like "always returning arguments along with intended result" or FP's "assuming no side effects" would be expected.

For example, a field of study which has an answer to questions like what functional addition, subtraction, multiplication etc. are with useful manipulation rules(I'm aware the starting point would be totally arbitrary/implementation-defined).

File: structheader.png (86KB, 1025x964px) Image search: [Google]

86KB, 1025x964px

A parse structure which contains options for parsing and some other data, with a structure for strings. The string structure holds present string count (for iterative/decrementive storage) and some other info, besides the special string pointer array. The special string structure holds the pointer to a normal string, it's place in the array, and recently, it's length. This way when I stop using a string, I can use the storage+index to deallocate all the data. If I create a pointer and run the creation function, it digs down and it's all safety checked and with debug messages. I've been at chapter 6 in the C 2nd book, and am rebuilding all my sort and loading functions for my structures. Since loading files is easy and my structures I dynamically allocated, it will be easier to test edge-cases now, and add more useful features later.

File: setactive..png (31KB, 658x692px) Image search: [Google]

31KB, 658x692px

>>62293709
>right click
>set active

I don't see it

>>62293871
They've changed the name, because of course they have. "Set as StartUp Project"

I'll post this here as well (from /fglt/):

I'm trying to get the position of a cursor in C/C++. The only way of doing this (without implementing some sort of logic to track the cursor position) is by using control sequence DSR as defined in ISO/IEC 6429 (see ECMA 48) with an argument of 6.

std::cout << "\x1b\x5b6\x6e";

This will add the following CPR string to the std::cout and std::cin buffers, assuming the cursor is at row 32, column 6.

^[[32;6R

By using the termios.h library, I can disable the terminal-echo of characters written to std::cout and I can make std::cin non-blocking. However, there's a problem.
If any input is written to the std::cin buffer (before or after it is made non-blocking) before or after the CPR is written to the std::cin buffer, it becomes difficult to parse, especially because I DON'T want the CPR in the std::cin buffer and I DON'T want to get rid of any user input (since that is up to the client to handle).
Here's a possible (partial) solution:
Swap the old std::cin buffer with another stream's buffer, *, request the CPR, *, put the old buffer back; parse the other stream's buffer.
But, at the positions marked by an *, the user may still pollute the stream with input and it will be lost even if the CPR is correctly parsed within the buffer. It's worth mentioning that I know ncurses exists, but I'm trying to solve this without using ncurses.

>tl;dr
How do I redirect the CPR generated by a DSR to a stream other than std::cin? Are there alternate ways of obtaining the cursor position without tracking it since the application began?

File: pvm.png (73KB, 1195x895px) Image search: [Google]

73KB, 1195x895px

A stack-based virtual machine. It can be programmed directly with a (somewhat) Forth-like language, but it's mainly supposed to serve as a target for higher level scripting languages, the main goal being interoperatiblity between them. The stack is a pretty comfy way of pushing data across language barriers, and the VM provides a simple but highly customizable object system (through meta-objects), so it's possible for each language to set it up differently but still directly share objects.

>>62293887
Thanks. Other random questions...

In the drop down menu, I have an option to build either the solution, or the project I'm currently working on.

Why would I not want to always build the solution? Or vice versa? I mean what's the incentive or benefit for picking one over the other, because there's gotta be one or both wouldn't be offered.

File: test2a.png (96KB, 967x442px) Image search: [Google]

96KB, 967x442px

>>62293887
also so I opened up the .vcxproj file in a text editor and I changed the name of the file in the text editor, the file name did not change in file explorer. And then so I figured, hey, I can just change it in file explorer- so I did, but then VS was still looking for the old project file (test2, not test2a) and I had to cut the project and reload the new one entirely.

Do projects just not support renaming? I mean that's fine it's just the first time I've ever seen a file I can't rename after creation...didn't know that existed lol

Why should anyone use linked lists over array/vector of structs?

File: 1498896707469.png (81KB, 240x240px) Image search: [Google]

81KB, 240x240px

r8 my code gee

https://hastebin.com/rukohavade.sm

>>62294216
constant time insertion & removal

>>62294224
unreadable

>>62294241
So should one use std::list on that scenario?

>>62294241
except the indirection for other uses often outweighs the small benefits of using list. look at the benchmarks. it's better the vast majority of the time to use contiguous memory.

>>62294251
sure

>>62293919

You can't use the built in C++ standard methods to do what you want via peek and get? I mean, you can probably use ignore too.

http://en.cppreference.com/w/cpp/io/basic_istream

In any case, if you really want to swap streams, there is a swap function for cin. Look through and see how you want to do what you want.

>>62294293
Read what I wrote again, you didn't understand it.

>>62294275
yes, linked lists have a rather small niche use
in production code it's unlikely the average programmer will ever want to use a linked list over a vector

File: DELETE THIS.png (3KB, 248x17px) Image search: [Google]

3KB, 248x17px

>>62288936
I know money should never be represented as a float, but is the ideal solution to store the dollar amount and the cent as 2 separate integers?

I think I should start writing blogs about my frustration while learning programming.

It's okay to have the students go through the basics of linked lists by making them implement them.
However it's also practical to introduce them to std::list because that's what they are going to be using most of the time


There are no good C++ books

>>62294313
store the amount of money in cents

>>62294321
the entire thing?

File: 16996520_110184319481004_1175464746939712463_n.jpg (51KB, 800x400px) Image search: [Google]

51KB, 800x400px

>>62294243
understandable, how about now?

>>62294330
yep, only requires 1 integer.

>>62294243
>>62294333
fugg forgot link :DDD
https://hastebin.com/setiwibogo.sm

>>62294224
>using the cuck license
trash

>>62294315
going over list implementations is good because it gives the beginner some experience working with pointers. but std::list is a terrible data structure (not by virtue of being in the standard library, but by being a list) for the vast majority of purposes. did you have a typo?
>there are no good c++ books
i can name at least ten

>>62294321
>not storing it in hay pennies

>>62294352
blue board please delete image thankings

>>62293919
just use curses

>>62294377
>It's worth mentioning that I know ncurses exists, but I'm trying to solve this without using ncurses.
I thought it would be redundant to say the same for curses, but I guess not.

>>62294352
BSD/MIT are the cuck licenses.

>>62294367
Yes most of the times you would be using vectors instead

>>62294352
Fuck off, /pol/fag. Stop projecting your fetishes.

>>62294352
The only reason why /pol/ talks so much about cucks is because they fantasize them being cucked.

>>62294373
ok
>>62294390
yes
>>62294409
>>62294397
not /pol/ at all

>>62294387
bruh, I read it
I'm telling you to use it anyway

>>62294251
Depends. Some std::list implementations have a horribly fragmented memory footprint. If you don't like that, it's not a big chunk of work to make your own implementation with an underlying object pool structure.

>>62294409
They're still butthurt after what >who? did to them.

>>62294422
thanks

>>62294425
>i won't use X
>"use X"
nice
i've used it before, and i'll likely use it again, but i am doing something different just for fun.

>>62294430
even in those instances, std::list can be shown to have poorer performance than std::vector in the vast majority of use cases, even when the "muh constant time insertion" sounds smart.

>>62294059
did you change it in the sln too?

>>62294482
but there's no clean solution outside of curses
unless you want to forget about linux and run your program in the windows terminal? because the windows API has what you want

>>62288936
I'm not a code monkey. Everything except MATLAB is for code monkeys.

>>62294306

I know you're asking for an alternative method to do your position tracking by not wanting to parse the damn stream but I was nudging you to reconsider because it's really not that hard with istream but the straightforward answer to your question is that without any of the limitations you already put on your damn terminal application with no logic tracking the position which you mentioned already. The CPR is sent the way it is because that was how it was done way back then as a legacy behavior.

If you want to get by without curses or ncurses, that's fine but you have effectively reached the limit on work you can avoid doing with your own stubborness.

>>62294603

*you mentioned already, it is impossible.

File: 2017-09-06-224140_1920x1080_scrot.png (196KB, 1920x1080px) Image search: [Google]

196KB, 1920x1080px

we're getting there i suppose

File: 1490484391641.png (68KB, 400x400px) Image search: [Google]

68KB, 400x400px

Why aren't you using Haskell /g/?

File: turtle.png (102KB, 600x400px) Image search: [Google]

102KB, 600x400px

>>62289398
Why use Python when you can use the best tool for the job, Logo?
http://www.calormen.com/jslogo/
https://turtleacademy.com/

File: file.png (62KB, 674x565px) Image search: [Google]

62KB, 674x565px

>>62294847
i am though

File: 1490484251913.jpg (32KB, 358x344px) Image search: [Google]

32KB, 358x344px

>>62295001
nice

File: 1497595537559.png (91KB, 347x259px) Image search: [Google]

91KB, 347x259px

>>62294847

>>62294847
>Haskell
Static typing and immutability suck for most of what I do.

>>62295087
>immutability
understandable, though most of it gets optimized away by the compiler
>static typing
now anon, there's never a bad type for static typing

>>62288936
Why do non-programmers have such a hard time with lain?

>>62289726
haha, no thread would be complete without haskell zealots trying to redefine language to make haskell """pure""".

If you're not being intentionally comedic here, this is a stupid argument that's been going on for at least half a decade on 4chan alone.

>>62295087
how come?

What's a good Unity3D intermediate course? Everything is aimed at beginners.

File: babfd6e9dbea631a298430751261a86d.jpg (175KB, 1620x1884px) Image search: [Google]

175KB, 1620x1884px

Why is Cirno such a dumbass?

File: 1503015221165.jpg (68KB, 642x640px) Image search: [Google]

68KB, 642x640px

>>62295198
wtf leave cirno alone

>>62295127
>most of it gets optimized away by the compiler
That may or may not be the case, but it gets in my way on the semantic level much of the time because I'm a game programmer.

>there's never a bad type for static typing
Again, it just gets in my way.

non-determinism is a jewish lie invented by the computer design ellite to keep us from understanding full semantics of properly written parallelly data accessing programs this is why such concepts locking threading mutexes etc are needlessly not formed of elementary CONCURRENCE-AWARE FUNCTION COMPOSITION programming of types by which the means to combinations of possible different functions result in one single well semantics function fucK NON-DETERMINISM fuck ((((((MULTITHREADING)))))) it is an inherently flawed model coupling implementation to function and as such this proves the determinate superiority of functional composition IN A CONCURRENCE-AWARE FASHION in the fucking YEAR 2017 where fucking dealing with non-determinism retarded retardedjewish lies of ((((MEMORY BARRIERS)))) just keeps us from expressing the true intent of our algorithms

>>62295240
cool

File: 1491273405568.gif (456KB, 480x360px) Image search: [Google]

456KB, 480x360px

>>62295226
B A K A

File: monads.png (42KB, 1081x702px) Image search: [Google]

42KB, 1081x702px

>>62295240
attached find a copy of the explanation i have provided for the generalised theory of usage of monad-like constructs in the design of a CONCURRENCE-AWARE FUNCTION COMPOSITION programming language specification to which hopefully the world is heading and will finally permit us to express the true intent of our algorithms in an easily understandable DETERMINISTIC manner which uncouples the behaviour and implementation of said aforementioned constructs in a clean way while incurring no performance penalty for the exploitation of several threads of execution or other such similar assured contexts in which the reasoning of humans breaks down but which is KEPT SO by the manufacturers and corporations that work towards our suppression for the sale of closed source properitary software which only they understand in full detail and hence prohibit us from using the our own machines paid by for us with conspiracy lies

>>62295240
>>62295289
do it you absolute madman

give me ideas for a game

File: 15f30bfd28c4ad180064c8caada44cbfebf53250161cd60d0b110e604ecd4935.png (228KB, 466x552px) Image search: [Google]

228KB, 466x552px

Is there a way to make my own "unique" looking shell sort code without following other pseudo code, or is it stupid to try and I should just follow something someone else has written down because it would end up looking the same anyways?

>>62295310
Ballistics

>>62295239
lua?

>>62295310
>ideas for a game
Turtle towers.

>>62295345
A custom-built lisp.

>>62295333
>>62295351
how about something in the theme of mentall illness

>>62295367
A game with a mentally ill protagonist launching turtles at enemies.

Trying to make a python script that reads txt files and tells me if certain words are in there more than a number of times. Any hints?

>>62295393
yeah, use grep instead

>>62291408
yes
i did

>>62295310
i had an idea for a game where you all kick each other in the nuts, but you only get to do so if you were kicked in the nuts already. When you log into the game, there's a small chance you'll get 3 free kicks, which is what starts the nut kicking brigade. It was going to be top down or 3/4 view. Probably 3/4 for better sprites.

File: Screenshot_20170907_171040.png (168KB, 1920x1080px) Image search: [Google]

168KB, 1920x1080px

So about friend functions, I find the way of defining the prototype inside the class and having to define the function outside the class goes against the DRY principle. For example what if I change the friend function? I have to repeat the process of defining the prototype inside of every classes. This can lead to bugs and frustrations.

>>62292272
I think nim can do that
Also LISP

>>62289761
A language can be purely functional while compiling down to stateful machine code.

The beauty of Haskell is that you can express stateful operations in a stateless language.

This might be a hard concept for your pea brain to understand, so I'll use a silly but simple analogy. Just because English doesn't contain Chinese characters doesn't mean we can't use different words to describe the same things as Chinese.

>>62295240
Did you read you SICP today?

File: Screen Shot 2017-09-07 at 2.20.34 AM.png (210KB, 1636x1034px) Image search: [Google]

210KB, 1636x1034px

>>62288936
Writing a shell script that creates a file with a basic template for specific programming languages. Want to get to the point to where after it's created it automatically opens, but I'm a brainlet and can't figure it out.

>>62295583
Wanna build a text editor instead?

>>62295503
why not make it inline?
..and why even make it a friend member? when it could easily be an class method?

but changing the prototype of a function will break the ABI, defining it outside of the class (and in the source part of the translation unit) lets you change the implementation..

and if you did want to change the signature, why not just provide an overload and/or type alias?

>>62295503
If you change the signature you're going to have to change the call sites anyway, that will never be un-dry because types are incommensurable.

The separation of the signature decl from the implementation is mostly to appease the linker.

>>62295522
>This might be a hard concept for your pea brain to understand, so I'll use a silly but simple analogy. Just because English doesn't contain Chinese characters doesn't mean we can't use different words to describe the same things as Chinese.
If you claim to be speaking Chinese but English-only speakers can understand you then your claim comes into doubt. A better analogy for why people see a program that's producing output to the OS as part of its execution and wonder if it might not be pure.

>>62295583
Save path to a variable for better handling (shorter commands, reused strings, etc..) and add '&& [editor] $file' after you create it.
'command1 && command2' executes command2 after command1 is successfully run.

Otherwise rewrite your script a bit, create variable with file+path, create in in case statement and open it after (if same editor for all), makes it shorter and easier to add new stuff.

>>62295503
>using namespace std;

You disgust me.

>>62295679
Thanks senpai.

>>62295680
std::this std::that std::oh std::fuck std::my std::shit std::up

>>62295680
idiot.

>>62295679
Additionally:
Save templates to variables before, makes script more readable.

Other ideas:
Make dir with template.[extension] and have your script copy the right template to new file based on extension.

>>62295709
>t. Brainlet

>>62295719
For >>62295583

new thread when

>>62295818
After the bump limit, you stupid fuck.

>>62295839
please sign my petition to move the bump limit on /g/ back to 300

There is no such things as bump limits

Dear /g/

I'm a professional software developer and I make absurd amounts of money, live in relative upper middle class luxury for like about 2 hours of real work a day. And yet I want to kill myself, it feels too easy, no attainment of wealth or academic merit has ever felt satisfying because it's just so easy. I was given all these things some people seem to want but I'm not entertained by them because they didn't require any investment at all from me.

Mostly I just want to drink and listen to alan watts tapes on the beach on the beach. I go to hawaii a few times a year to do this because it's the only way I can tolerate my absurd existence.

What do?

>>62295849
/dpt/ must be a rolling thread

>>62295976
grow up and stop dreaming, or just do it already

>>62295976
Work on personal projects in your free time. Finding something you're passionate about is probably the most important goal in life.
You're at a great starting point: lots of money and free time. Time to put them to good use and get to real work by finding a challenge for yourself.

>>62296010
do what though? All I know is programming and some CS. Whenever I think about doing something else I realize everything people are willing to pay me for is stupid and if I'm going to do stupid things with my life I might as well do the one I'm good at for some reason.

It's like my ma always said, "if you gotta eat a shit sandwich, take big bites"

>>62296041
>do what

That anon meant "kill yourself" by that. Anon is very funny.

File: 1504299434792.gif (1MB, 270x235px) Image search: [Google]

1MB, 270x235px

Is Python the BASIC of the XX century?

New thread:
>>62296062
>>62296062
>>62296062

Anyone tried free pascal and lazarus? I'm looking for a less bloated way to have a multiplaform gui than java or qt and it seem to fit the bill.

>>62296069
Qt is fucking amazing, tho.

>>62296076
I know, just looking for an alternative when i don't even need a third of the features or have to produce a quick working prototype.

>>62296083
DELET THIS

>>62296089
There's Qt Lite for debloating. Might want to check it out:
https://www.youtube.com/watch?v=p0oZWC3TcRo

>>62291685
Can you know in advance what you need to do with each query?
You could write a program/script to do all the waiting for you.

>>62295731
That worked perfectly dude. What if I want the file to open in visual studio code, for instance, instead of emacs?

>>62296134
'[name of visual studio code executable] $file'

>>62291936
Check out boost multi index map if you're using C++ (and boost)

>>62288936
Solving a USACO problem about Wormholes.