/dpt/ - Daily Programming Thread

First for racket macros are awesome

How much math do you need to be a competent programmer?

>>56122356
Learning Python I just typed Hello world

>>56122407
Primary school for web and java,c# crud

>>56122468

  File "<stdin>", line 1
    hello world
              ^
SyntaxError: invalid syntax

why is programming so hard

>>56122407
About 12.

>>56122407
You need to be good at it but you don't need to have studied a lot of it, unless you do something like graphics programming, computer vision, and other research-y stuff. If you want to make a GUI program with pictures of all of your dildos, then you're fine.

>>56122502
Even for researchy stuff you dont need too much math, linalg 1 maybe 2 some calculus and maybe diff eqs...

>>56122579
That's a lot of math for someone that posts this >>56122410

>>56122183
Just use IO

>>56122312
you weren't kidding

>>56122356
Anime is trash

>>56122608
Well i do but wrapping stuff in monads and using fmap and lift for basic stuff is annoying... I know how to do it but that doesnt mean i like it.

>>56122183
Just use Ocaml if you are annoyed about purity

>>56122632
Lisp is my true love

How's OCaml, never seen any code... Does it have libraries?

>>56122631
I don't really understand the issue you have

Have you guys ever seen people on /g/ outside of /dpt/ talking about programming? It's really fucking awful.

>>56122189

>>56122628
you're trash

>>56122650
No practical issue, i can write programs in haskell haha, i just dont like it. I love free form languages where i can do whatever i please

>>56120504
What a fucking coincidence, you're retarded, AND using java

So guys, I am bored at work for the next 6 weeks (I study and work in semester pauses, but there is 0 for me to do ATM and no one even cares).

What the heck can I do?

Apparently we will get some new frameworks that MIGHT use same scala things.
So I am reading into scala right now. It's meh.

Anything fun that I might develop in scala do really get the hang of it?
Or in C/C++/Java. Anything really, I don't want to read documentation 8 hours a day.

>>56122651
all the time

i went in there and called the OP a faggot and he basically claimed that knowing hexadecimal is completely unnecessary and irrelevant

>>56122663
Do you like python> That's how it feels to me, despite being pretty messy

>>56122677
I hate python more than java. I like C, all lisps, kinda C++, x86 asm, even Scala

>>56122640
well if you care about libraries as welll, then use F#

all the .NET libraries thanks to seamless interop and P/Invoke for C/C++ and you get cool parallel shit

>>56122676
I think 2 dpt's ago there was someone that didn't know how to convert from hex to binary and kept complaining ("Sorry that's gibberish to me :\")

I don't get how retards get into programming. Is it just that they want to make games and websites?

>>56122683
You're pretty gay dude. It seems like /dpt/ formed that opinion for you.
>all lisps
Crazy!!!!!

>>56122677
Its personal taste nothing wrong with language itself

>>56122702
dpt pro lisp? Ha? They always spout how haskell is good... I used lisp even before i knew about 4chan...

>>56122690
they get told that computers are a lucrative job market and they want a piece
they don't actually like computers or programming tho, they just want a paycheck

>>56122651
Now who the fuck was the guy that made the shitty C prototype and tried defending it?
I bet he posts in here too.

>>56122782
What do you want?`

>>56122651
God there are people in that thread saying math has 0 practical uses

>>56122812
I sure as hell hope you didn't complain about "people that talk about programming outside of this thread".

What's pull request etiquette on Github? I forked and wrote an improvement to a small library. Should I just submit a pull request out of the blue or contact them first?

>>56122828
I'm not sure what you're talking about.

>>56122836
You can expect to have your pull request denies for such reasons such as "this is outside of scope" or "i don't like your coding style".
Then a few days later, they'll push a commit that adds in almost all of your changes, but you won't get any credit.

>>56122825
Yea, I'm right here. Uni math is useless in every day tasks.

Unless you do something very specific you will never use it again.

By the time you might actually encounter something you leaned in university's first semester (which is where they filter with abstract maths) you will have forgotten it anyways and need to relearn it.

Again: I am not saying that the fact there is abstract maths in university is bad. It has its use and it makes sense that students have to learn it.
But that is more for the fact that they should be able to learn such complex topics and less because they will actually use them later on.

>>56122842
This >>56122651

>>56122880
that wasnt me

>>56122855
Dude in that thread gave great example with determining when 2 cars meet. Mathematics is used for abstracting and simplifying problems. Also if you want to do anything but crud apps and front end you need math.

>>56122912
So what? Even if you never had maths in your entire life, you would still come up with a solution by googeling and understanding the maths behind it within minutes.
And you can do that because you learned how to understand mathematical matters in university.
But you don't actually use things you learned in university.

Just as I said. You should read my post again, you are missing the point.

math is a meme

any necessary math you need can be googled and understood well enough to implement in software

you don't need university math to program, proofs are mindless busywork, just like doing vector calculations by hand

>>56122982
>>56123007

How do you suggest implementing anything more complex without understanding math first? Do you think someone without math skills could implement perlin noise? dft? fft?
>use libs, what if library implementation isnt specific enough for your use case?

>>56122663
Well it's more like a type error, but i guess that's fine

>>56123020
leave that to mathematicians

>>56123020
If you were to work in a field that requires you to apply higher level maths, you would be interested in higher level maths and you would learn it during the process you work in that job.

You still wouldn't use what you leaned in university. You would learn what you need by studying it yourself of being educated through your job.

>>56123032
>>56123030

Yes you dont need math for pajeet level stuff. agreed. But then you write pajeet level code as well

>>56122828
No, did you read my post? I complained about how people on /g/ outside of /dpt/ usually talk about programming.

>>56122782
You're retarded. He asked the anon to implement a hex to decimal function, and anon posted "I googled it and here it is" and just posted calling a standard method to do it, it seems. He posted a prototype to clarify the question. Complaining about a "shitty" prototype in this context is retarded. I bet you're the guy he was talking to.

Last one for tonight, it's 4AM. But this is fun.

tupleIze :: [Double] -> [(Double, Double, Double)]
tupleIze [] = []
tupleIze (x:(y:(z:(as)))) = (x, y, z) : tupleIze as

averageTuple :: (Double, Double, Double) -> Double
averageTuple (x, y, z) = (x + y + z) / 3

doubleAvgs :: [(Double, Double, Double)] -> [((Double, Double, Double), Double)]
doubleAvgs n = [(t, (averageTuple t)) | t <- n]

averageShow :: [((Double, Double, Double), Double)] -> [Char]
averageShow [] = ""
averageShow (((x, y, z), a): xs) = show x ++ ", " ++ show y ++ ", " ++ show z ++ ": " ++ show a ++ "\n" ++ averageShow xs

main :: IO ()
main = do
    putStrLn $  averageShow $ actualAvgs $ tupleIze [1.2, 5.6, 9.8, 10, 4, 7, 15, 12, 14]

Output:

1.2, 5.6, 9.8: 5.533333333333334
10.0, 4.0, 7.0: 7.0
15.0, 12.0, 14.0: 13.666666666666666

Neat!

>>56123063
forgot to replace actualAvs with doubleAvgs, changed that before recompiling, but you get the jist

>>56123073
why not just map + and div by len? That seems overly complicated

>>56123081
Because I'm learning :*)

>>56123084
Have fun, code is fine

>>56123049
The prototype was retarded and I named numerous arguments as to why it was retarded.

What the fuck is your point even? You mean I cannot complain about the prototype because that was not the purpose of the his post? What the fuck

>>56123081
Also, I'm averaging every three doubles in the list, not just averaging the whole list

>>56123063
Anon
x:(y:(z:(as
x:y:z:as

>>56123097
Ugh, C - ugly.

>>56123047
You're too dumb to understand the point. What a retarded dicisission.

If you want to understand the point, redraft my posts and use your brain this time.
I never claimed that you "don't need maths" btw nor did I say that learning maths at university is stupid.

>>56123106
Thanks! Page I'm reading used x:(y:(zs)) so I thought maybe I needed it.

>>56123107
Wasn't me. Do you have any actual point or what?

>>56123118
Then what the hell are you arguing for? And ofc you use math you learned at uni, if you didnt learn it you wouldnt have those math skills you mentioned in first post

>>56123124
Ugh.

>>56123063
Please don't use tuples for no reason.

import Data.List (intercalate)
main      = mapM_ (putStrLn . toLine) (chunk 3 nums)
toLine xs = intercalate ", " (map show xs) ++ ": " ++ show (mean xs)
mean xs   = sum xs / fromIntegral (length xs)
chunk n   = takeWhile (not . null) . map (take n) . iterate (drop n)
nums      = [1.2, 5.6, 9.8, 10, 4, 7, 15, 12, 14]

>>56123211
to me, his version is more readable

>>56123229
It's okay if you aren't very good at Haskell yet, you'll learn!

But seriously, there's no excuse for manually recurring over lists when map exists.

File: Haskell.png (75KB, 1549x683px) Image search: [Google]

75KB, 1549x683px

>his language can't even manipulate types
where my C++ template bullshit bros at

>>56123134
No you don't. You use your learning skills you acquired.
Not the maths you learned.

>>56123258
How do you suggest learning math skills without learning math? Also you do use linalg 1,2 calculus, statics probability discrete math...

>>56123211
The point wasn't to just get the average, but to display them how I did in the output, too
I did it without displaying it the first time and of course it was much smaller, didn't have tuples, etc.

>>56123252
whoops, with TypeInType and DataKinds you can actually just do

type List = [Type]
and then define the type families using '[] and :

>>56123265
Sigh. The point still didn't come across.

Forget it. I have to work now.

>>56123299
The code I wrote gives identical output as yours.

File: LineMayhemTVBanner.png (81KB, 1280x720px) Image search: [Google]

81KB, 1280x720px

>>56122356
I finished an Android game that I built directly onto Android (no Unity, LibGDX etc) two days ago so I've been shilling and fixing a couple of bugs.
It's called Line Mayhem guise, it's simple but challenging, and it's not a flappy bird clone, it actually has it's own concept and I've received mostly good reviews.

you should check it out https://play.google.com/store/apps/details?id=com.MightyBarbet.LineMayhem

pls check it out

File: magic.png (36KB, 707x394px) Image search: [Google]

36KB, 707x394px

>

File: bvcbc.jpg (735KB, 2260x2056px) Image search: [Google]

735KB, 2260x2056px

More Usenet anime indexing. Dropped the project for a while.
I think I get 5 people a day actually using the website, which is enough for me.

Could anyone recommend a good beginners book on making android apps? I have done some java programming.

>>56123479
wait what site? Link?

>>56123020
>Do you think someone without math skills could implement perlin noise? dft? fft?
Of course.

>>56123692
Right i was bit too permissive with that, implement and understand.

>>56123704
Yes, if you googled it, a person without math skills should be able to implement and understand it after thinking about it enough.

>>56123735
To add, if you can't do that, then you probably shouldn't be programming in the first place.

File: 1464368071633.jpg (23KB, 490x480px) Image search: [Google]

23KB, 490x480px

>>56123252
>where my C++ template bullshit bros at
Mah nigga
http://pastebin.com/2uH3GZDN

>mfw C++ implements a dynamically typed language at compile time

>>56123735
Well for understanding perlin you would need to understand linalg and numeric, while for fft you need complex analasys. While you could learn that in principle i doubt you would in practice.

>>56123747
>While you could learn that in principle i doubt you would in practice.
So what, it's not impossible.
I taught myself linear algebra in a somewhat short time in order to do 3D transformations, and I consider myself shit with math.

File: runtime get out.png (22KB, 1161x407px) Image search: [Google]

22KB, 1161x407px

>>56123746
>lain
excellent taste

>>56123793
Are those compile time monads in C++?

>>56123806
no, >>= is just (the equivalent of what is now) std::apply

but you could do that too
bind f for lists is just concat . fmap f

>>56123786
Then we are in agreement. Stil you need math for cs. For coding no while it would have been easier for you if you learned that at uni

Making 'botnet'.
When is the best time to screenshot? What trigger should I use?
A mouse click? I have it on a timer right now.

>>56124028
Well you probably want to take shots only of web browser since thats where interesting stuff goes, so time after browser launch?

>>56124045
Sure, maybe.
I screenshot when switching active windows also, I'm not sure how to get relevant information after that, though. /maybe/ title of window changes... But, it's likely going to be "facebook" all of the time, right?
hmm, maybe trigger for return also..
What else?

>>56124028
routine screenshots every ...15 seconds? or maybe just have it stream the desktop

>>56123669
anizb dot org

>>56123479
make it wider
there's too much blank space

File: output.webm (2MB, 260x462px) Image search: [Google]

2MB, 260x462px

Muh space shooter, now with Ads! (AdMob, iAd is kill)
Will you click on my ads /g/??

>>56122483
Welp

>>56124097
Depends on which os you are targeting but to get process name on winblows you can use powershell get-process -DisplayName "firefox", so there must be something in ,NET.

Yea its most likey to be facebook but you dont do banking outside of browser or anything of real interest most of the time

>>56124255
It's shit

>>56122825
I can literally feel the retardation in that thread

>>56124240
I think so too, but I think the off-white background makes it look a lot worse than it is.
I'll make it a bit bigger.

>>56124255
You've betrayed us all
I trusted you

>>56124255
Imagine having adds on your desktop games? Why is this common practice on android?

Fdroid is only usable repo...

>>56124319
b-but it will still be open souce anon!

>>56124331
> desktop games
wew lad, if you don't give me money, I'll give you ads.

>>56123326
Go write that pajeet code, front end boi

>>56124473
Ever heard of free game on windows thats not farmbille or some shit having adds? If you want money charge for your product

File: 200w.gif (584KB, 200x115px) Image search: [Google]

584KB, 200x115px

>>56123793
>nomenclature

>>56124538
but my game is a garbage space shooter that I will probably never update

I will also allow the user to hide the ads for $0.99 and sell shitty items

>>56124571
....
....
Im speechless

>>56124592
I would definitely download GTA V if it was free but with small ads on a few menus

>>56124623
But you dont have anything of bearly the same quality on android and everything is either paid or full of adds. I have either to torrent or use fdroid...

>>56124674
And i dont mean graphics i mean content

>>56124258
>so there must be something in ,NET

Process.GetProcessesByName(processname);

You can just pass it "firefox" and it hooks directly into the process.

I use this to automate shitposting.

>>56124571

Or the user could hide the ads with Ad Away and you'll make nothing on them.

>>56124255
Your game is shit.
It looks like practically no effort went into making your game.
I've literally created better shoot-em-ups for shitty school projects.
Your game is not Touhou.

>>56124674
> make shitty paid app
> no one buys it
or
> make shitty free app with ads
> hundreds of thousands of downloads
> throw ads on their faces
> get money for my work
> if you don't like ads buy the fucking app
What's wrong with this?

>>56124698
Any tips on that?

>>56124729
Nothing bad for you. I wouldnt download it tho. Adds are no go for me, nor would i buy, ever. I saw enough of fhem on tv nomway ill bring them to tablet or laptop

>>56122483
What did you fuck up on a
>print "Hello World"
sentence??

File: autopost.webm (319KB, 1740x462px) Image search: [Google]

319KB, 1740x462px

>>56124733
What do you want to know specifically?

>this is an automated shitpost, please disregard

You just need

System.Diagnostics

and

user32.dll

:

[DllImport("user32.dll")]
static extern bool ShowWindow(IntPtr hWnd, int nCmdShow);

>>56124815
Ooh thats neat thanks!

>>56124722
My app has an ad away detector, whenever the ad fails to load, the app freezes and It takes a picture of you and upload it to a jewish database so one day you will get blackmailed

Can anyone recommend me videos of someone coding a project from start to finish, every single line visible and explaining along the way? For JavaScript.

Even if it's hours long, I think I learn best by shadowing what someone is doing even if I don't understand it.

>>56124848
Keep in mind that those are literal keypresses, so you could do commands or shortcuts if you wanted to roll your own macro of sorts.

>>56124877
Any tutorials perhaps? I might give windows another go with this

>>56124875
I'm not memeing you: books are infinitely better for this. The code is all there, line by line, with explanations.

Almost all books ramp up apps as the book goes on and culminates in a single somewhat complex application with all the source code laid out in front of you.

Plus, you don't have to keep tracking the video to catch exactly what was written.

Alternatively, use something like PluralSight.

>>56124897
Sounds good, can you recommend me any books like this?

>>56124892
What specifically are you looking to do? Windows DLL calls, or C# in general? You don't need to be on Windows to write things in C#.

There's tons of shit you can do once you start dipping into Windows dlls.

Here's a simple keylogger done in like 3 minutes:
https://www.youtube.com/watch?v=-wQRyi24WXE

The guy displays it on a form, but you could obviously record that to a file, and send the information over the web to a Google Drive API or whatever. You could write this to work on Windows, Linux, or OSX, but I doubt it'd get past any AV as-is.

>>56124940
Well catching keystrokes so i can make keybindings. Windows dll calls c# should be no problem

why does this return

Symbol's value as variable is void: list-of-names

(setq list-of-names '("Sarah" "Chloe" "Mathilde"))
(push "Steph" list-of-names)

>>56124940
To add to>>56124970
Mostly dll calls since im not sure what you can do i dont know what exactly im looking for. Stuff like automating shitposting, macros, keylogging etc

>>56122356
Why do you degenerates like anime? It encourages whites to racemix with japanese as a form of white genocide. Plus weebs are NEET cucks

>>56125040
cuck

>>56125045
Only cucks are race mixing weebs that hate their own race

>>56125050
cuck

>>56125077
Projecting

>>56125040
>He can't reverse a binary search tree
LOL!

@56125089

so true! 100

>>56125089

var binarySearchTree = new SortedDictionary<int, object>();
binarySearchTree.Reverse();

>>56125109
write it in C, champ

>>56125120
No.

>>56125131
kek

get outta here, pleb

>>56125136
>he can't do thing
>oh shit he did thing, better add requirements to sate my autism

>>56125131
>He calls others degenerates when he can't reverse a binary search tree in C
LOL!

>>56125226
I didn't call anyone a degenerate, faggot. That was a different anon.

Anime is fine, and C is a waste of time for the vast majority of programmers.

>>56125021
>>56124970
>>56124940
doing it all wrong, get a WDK version
https://github.com/jasonpang/Interceptor

File: shhhh.jpg (278KB, 800x600px) Image search: [Google]

278KB, 800x600px

I have just about finished making my own implementation of a data analysis tool developed and maintained by a certain research institute. It took me about and week and I did it because:

1. Their tool has a few bugs that can sometimes (often) ruin results.
2. Their code is convoluted, unreadable Perl spaghetti code.
3. Their tool takes 30-40 minutes to process a single sample and there can be hundreds or even thousands of those.

I'm still doing some tweaks but my version of the tool doesn't seem to produce any errors and can process a sample in 20-30 seconds.

I managed that through the power of C, fuzzy hashes, and caching like a madman.

I smell a publication.

>tfw the Perl tool by the other group was developed and is maintained through a $1.5M NSF grant

>>56125872
*it took me about a week

File: 134257380020.jpg (34KB, 288x499px) Image search: [Google]

34KB, 288x499px

>People still fervently defend VB.NET and get offended when you tell them it's a language purely designed to bridge the gap between older developers and modern development

>>56125872
Also the line counts:

Their Perl abomination: 7500 lines
My beauty: 913 lines
(it's mostly Cython doing some parsing and basic processing wrapped around C doing the heavy lifting)

The functionality and results are identical except mine doesn't break on certain combinations of edge conditions (the aforementioned bugs).

I am trying to understand arrays.

package Examples;


public class Test{
public static int banana(int[] a){
    int kiwi = 1;
    int i = 0;
    while (i<a.length){
        kiwi = kiwi * a[i];
        i++;
    }
    return kiwi;
}

    public static void main(String[] args) {
       //int[] a =  banana([4] )
    }
}

What do I write into main, in order to make this value method work???

>>56126099
That function returns an int not array

>>56126099
int a = banana(an array)

>>56125891
Stop avatarfagging

>>56125109

The premise of this challenge is that you are using a BST without a builtin reverse method. You actually have to do the work yourself, rather than relying on the standard library. The idea is that if you can't do something as simple as reversing a BST, you can't be expected to handle more complicated tasks very well.

Assume we have an object representing a BST. It has three fields: left, right, and data. left and right are either undefined or another object of this sort. Data can be anything. If x is an object of this sort, write a function reverseBST(x) that returns a new object that is x, reversed.

File: hoerse.jpg (27KB, 221x294px) Image search: [Google]

27KB, 221x294px

Whati s the best way to delimit packets in a tcp stream?
Im currently using a sequence of 3 bytes which i check for every time data is received, but im worried some random data might trigger it down the line accidentally.

>>56126247
use msgpack

Hello. I'm hoping someone could give me some tips. There's a few things that I very much dislike about my RPG.
> One is that my methods call other methods. Would it be best to return a few of them instead of calling them? I fear of reaching the maximum recursion depth if I continue this way. Would this be efficient?

## old call
phase['create'].assign_class()
## new return
return phase['create'].assign_class()

> Two, I've gotten better but sometimes I still have trouble reusing logic. if you look at my method skills() you'll see this about five times.

if creation:
  phase['attributes'].skills(creation=True)
else:
  phase['attributes'].skills(creation=False)

So right now I consider it spaghetti code because it's not that great. The issue is mainly just my problem solving but I would really appreciate a few tips. Here's my RPG so far.
https://github.com/enigma424242/RPG/blob/master/textbasedrpgtest.py

>>56126224
>The premise of this challenge is that you are using a BST without a builtin reverse method. You actually have to do the work yourself, rather than relying on the standard library.
You're proposing an entirely different challenge.

The other anon simply mentioned reversing a binary search tree, without any other context.

Go fuck yourself, ponyfag.

File: 1453657295316.jpg (112KB, 1920x1080px) Image search: [Google]

112KB, 1920x1080px

>>56126432
>if b:
>.......(true)
>else:
>.......(false)

>>56126467
>Not even taking the time to read it.

>>56124875
This is one of the best examples I know of that:
http://www.nicofinder.net/watch/sm8517855
It helps to know Japanese though.

>>56126770
Oh wait I didn't see the "for Javascript" part.
The same guy did do an Othello in Javascript:
http://www.nicofinder.net/watch/sm8391299

File: 1441318392832.jpg (79KB, 600x720px) Image search: [Google]

79KB, 600x720px

I'm studying C and found a program that uses the

int argc, char *argv[]

part. I never used it before (my int main are always void) so my question is, is it important to learn how to use this shit or should I ignore it?

Here's a cute maki pic in advance.

>>56126843
yeah you should know how to use it. argc is the size of argv. argv is an array of "strings", which are your command line arguments given.

>>56126843
You should "learn" it
by "learn" I mean it will take you 10 seconds

argc is the number of elements in argv
argv is the arguments passed to the program via command line
argv[0] is the executable name and possibly path

Are you using python3?

File: 1469370134627.jpg (111KB, 436x626px) Image search: [Google]

111KB, 436x626px

>>56126900
>>56126935

So I've been programming wrong this whole time?!

>>56127142
It depends what you're writing, really

>>56126843
In standard C, there are two valid prototypes for main():

int main(void)

int main(int argc, char *argv[])

or anything that is equivalent to that.
If you are going to use command line arguments, use the second, otherwise use the first.
Also note that

void main()

is NOT valid, and is implementation specific.

>>56126770
>>56126826
Thanks, hopefully I can follow it without knowing Japanese!

>>56126935
You should also mention that argv[argc] == NULL and that argv[0] may be NULL or "" if the program name can not be found.

>>56126843
>>56126900
>>56126935

See how maki pic always results in helpful and useful advice. Maki forever!

>>56122468
u shoud have type "Why I am still living?" DESU

>>56127202
Or on obscure platforms

>>56127205
This desu des

>>56126944
If you're refering to
>>56126432
I'm using Python:2.7

>>56124872
same desu, but I also ask them to turn it off.

>>56127189

void main()

Not valid? Ha didnt know that, do you know in what way its implementation specific?

>>56127404
sorry didnt see void there thought it says

 int main()

>>56126089
>>56125872
>academic software

It's always horrible.

I've been self-studying Haskell for about 3 months (Don't know any other language yet) and I was wondering if you are able to generate graphics like this ones
https://www.youtube.com/watch?v=uQhTuRlWMxw
He is using python, here it's the code/library
https://github.com/3b1b/manim

>>56127517
there are quite a few Haskell graphics libraries

you can always use a FFI library to call Python from Haskell
https://john-millikin.com/articles/ride-the-snake/

>>56126432
> One is that my methods call other methods. Would it be best to return a few of them instead of calling them? I fear of reaching the maximum recursion depth if I continue this way. Would this be efficient?
You are WAAAAAY safe concerning this stack overflow thing. Trust me. So don't fix it if it ain't even broken.

Also I believe Pyshit doesn't do tail call elimination, so your proposed changes will do nothing. Did you come from Lua by any chance? What disaster brought you down into hell.py?

File: 1462122599124.jpg (108KB, 540x720px) Image search: [Google]

108KB, 540x720px

Every time i start with a project i have a detailed plan that i want to follow. But i always end up changing and changing to make it easier and more efficient and i have to rewrite almost everything each time. I end up spending weeks on writing different versions the same program over and over again. What does g do to prevent this? Should i just keep on going with my first design no matter what? Or should i just spend more time designing?

>>56127684
Use agile methods in your builds.

Get iteration 1 out and actually working before you go and rework UI and polish things that aren't quite great.

All of my code pushes are associated with tasks/user stories that are assigned to a particular iteration. If I don't like something, that's fine; just create a story for it and drag it to later iteration. As soon as the iteration 2 sprint actually starts, I'll re-visit it.

>>56127719
Sounds like a plan. Ill try this

>>56122483
stupid niggers

File: anal beads.png (38KB, 1476x616px) Image search: [Google]

38KB, 1476x616px

>>56127748
If you use VS, you can just use Team Services as version control + project management.

It's free for <5 person teams.

At the very least, Team Services is a free unlimited file storage; they haven't bothered to put a limit on it yet.

One guy has files being "version controlled" over 8 TB and he hasn't paid a dime.

File: McNulty.webm (3MB, 1280x720px) Image search: [Google]

3MB, 1280x720px

/dpt/-chan, daisuki~

Ask your much beloved programming literate anything (IAMA).

>>56125872
What the backstory of that pic ? is the dog dead ?

>>56122407
Programming is one of the most difficult branches of applied mathematics; the poorer mathematicians had better remain pure mathematicians. The easiest machine applications are the technical/scientific computations.

>>56128087
It's a dog with a toy raptor propt on top

Learning racket by learning and implementing neural network. Maybe taking new lang at a same time as new topic wasnt best idea but its character building excercise.

Writing a LLVM compiler for Andromeda in Agda.

How the fuck do you write good code in an OO language? We switched over to C# recently and I'm so lost. In C, I write the functions first, then if I notice variables get manipulated together frequently, I wrap them in a struct. In C#, everything needs to be part of a class, so I wind up with classes with embedded classes with embedded classes, with redundant information because >muh encapsulation

It's impossible to know the ideal way to organize code unless you've already written it, and with OOP, the amount of boilerplate makes refactoring a pain in the ass, so I just duct-tape things together instead of doing it "right".

>>56128138
Well OOP is kinda bloaty. And muh composition doesnt really work

Is it a bad idea to start learning from how to books from a bookstore?

>>56128138
You can use structs and unsafe memory management in C#, if that's your cup of tea.

Are you complaining about the policies your team is enforcing, or your lack of understanding of the language's features?

>>56128244
How do I write a function that isn't attached to a class? This is the number one issue. Most functions can be called from several classes and shouldn't mutate state, but making them all static just adds more boilerplate

>>56128233
yes

>>56122651
Have you guys ever seen people in /dpt/ talking about programming? It's really fucking awful.
They are always arguing about stupid, useless shit.

>>56128269
What is your intention?

You want publicly available functions that any piece of code in your project can call at any time?

Do you understand the static keyword and its implications on how the memory is allocated? You're coming from C, so you should have a basic understanding of stack and heap.

Here's a pretty good article that should answer some questions your having:
http://www.codeproject.com/Articles/15269/Static-Keyword-Demystified

>>56122912
>Uni math
>Dude in that thread gave great example with determining when 2 cars meet.
wut. you should have been taught this shit in HS

>>56125872
$10 says I can do better. What's the tool?

>>56128308
fug :DD

>>56128342
Yes. I understand what static is. I'm saying there's no way to have a freestanding function in C# that I know of, i.e., what all functions in C are. So literally every function has to be attached to a class, which is nonsensical from an organizational perspective. If I have a particular math function like add_or_subtract(int a, int b, bool flag) that either adds or subtracts based on a flag, I need to write a wrapper class and call it like wrapper.add_or_subtract(...). The add_or_subtract function doesn't need any state, so why should it have to be wrapped in a class? It makes no sense.

File: 1425864833588.jpg (566KB, 1204x1392px) Image search: [Google]

566KB, 1204x1392px

>>56122356

how can I make this chunking algorithm faster? it takes like 5 minutes to run through a large amount of points (called spawns in the code)

def getChunks(spawns):# a quick* chunking algorithm that chunks the spawns
# *it was quick for me to make at least
    iterations = 5   
    def chunks(spawns):
        maxDistance = .250
        foundIndexes = []
        output = []
        for i in range(0,len(spawns)-1):
            found = []
            for j in range(0,len(spawns)-1):
                skip = False
                if j not in foundIndexes:
                    d = getDistance(spawns[i]["latitude"],spawns[i]["longitude"],spawns[j]["latitude"],spawns[j]["longitude"])
                    if d < maxDistance:
                        found.append({"latitude":spawns[j]["latitude"],"longitude":spawns[j]["longitude"]})
                else:
                    skip = True
            if found:
                newPoint = avg(found)
                output.append({"latitude":newPoint[0],"longitude":newPoint[1]})
                foundIndexes.append(j)
            elif skip:
                output.append(i)
        return output

the gist is that it groups points that are within a certain distance into the average of all the points in the range. Then it runs the output back through the loop 5 times.

What kind of performance advantage would I see if I got rid of all the dictionaries? and what if I implemented it in C?

>>56127684
HNNG

>>56128087
>Programming is one of the most difficult branches of applied mathematics
lols

>>56128422
coming from C that can seem wrong, but is actually useful

if you have static class say MathUtils you know where to look for such math functions
it also reduces namespaces problems
all in all it's quite an elegant solution

>>56128440
Do it in Cython you idiot

>>56128440
>uses python complains its slow
You dont have to call len for each iteration of i.
Problem is probably in getDistance not in getChunks

>>56128422
As to your example of stateless, publicly accessible methods, it's merely a means of organization.

The function

AddOrSubtract()

is extremely vague and could have many different meanings.

Are you adding/subtracting numbers? Adding/subtracting items from an inventory system? From a checking acccount? etc.

Luckily, you can easily design your program to use this behavior that you desire.

Simply put all of your methods into a class called MyGodMethodsFuckState, make them all static, and put the following at the top of your code:

using static namespace.MyGodMethodsFuckState;

Now, you can call add of your methods without specifying a class name.

Again, there are many MANY other implications and uses for classes, static or otherwise, but for your specific issues, they're a means of organization, and specifying what flavor of

AddOrSubtract()

you desire.

>>56128493
len is O(1)

>>56128525
strlen() is O(N) dumbass

>>56128525
Well i didnt say that was main problem, even if its O(1) its (n * c) times more expensive than saying
foo < CONSTANT

>>56128493

All get Distance is is a bunch of math functions. I haven't gone through it very carefully yet, but I don't think it's worth the time to pick through it

def deg2rad(deg):
    return deg * (math.pi/180)

def getDistance(lat1,lon1,lat2,lon2):
    R = 6371; # Radius of the earth in km
    dLat = deg2rad(float(lat2)-float(lat1))  #deg2rad below
    dLon = deg2rad(lon2-lon1)
    a = ( 
    math.sin(dLat/2) * math.sin(dLat/2) +
    math.cos(deg2rad(lat1)) * math.cos(deg2rad(lat2)) * 
    math.sin(dLon/2) * math.sin(dLon/2) 
    )
    c = 2 * math.atan2(math.sqrt(a), math.sqrt(1-a))
    d = R * c #Distance in km
    return d

>>56128543
>strlen() is O(N) dumbass
It's actually not. In most higher level languages, including python and PHP, the length of a string is stored as an attribute of that string, and the length function just retrieves that value. You would be correct if strings were stored as nothing but arrays of characters that needed to be iterated through to compute the length, but this is not the case.

>>56128440
Try using a non-meme language

>>56122407
all you need to know to program is how to google effectively.

I feel like my job as a programmer is not to write code, but to search the web, parse and compile it to in a consistent format.

>>56128626
according to /g/ every language is a meme

>>56128440
>how do I make x faster in python
use numpy

>>56128644
We all start off with google, but after many years, if you're still looking at google to look up things you've done hundreds of times before, you're not a programmer, you just copy paste and are no better than an assembly line worker.

>>56122356
I need a good C++ book, something that covers C++14. Any anon want to point one out?

File: developing.jpg (87KB, 600x788px) Image search: [Google]

87KB, 600x788px

>>56128680

>>56128685
"Learn C++14 in 24 years"

>>56128685
What exactly is it that you want to learn from such book?
"covering C++ 14" could mean any number of things.

>>56128685
if you have 15' display or smaller any book will do

Trying to figure out how to create mipmaps.

Sure, I can average four pixels just fine. It's the alpha channel that gives me trouble. See, a while ago I had this incredible idea of creatively abusing the alpha channel to store non-alpha data, then use it to generate the actual alpha value in the fragment shader. It works fine as long as I don't use mipmaps. But things get really blocky and aliased without them.

I've tried to "intelligently" analyze the alpha values of the four pixels and generate a new value based on them, but all I get is ugly fringing and halo effects.

I could split the texture into two parts: GL_RGB for the mipmapped pixel data, and a special non-mipmapped GL_RED texture that contains the alpha data. It might work, but then I need two texture lookups per fragment. And some triangles are multi-textured, so it's four textures per fragment... all the texture switches and lookups will kill the performance.

Anyone have any creative ideas I could try?

Making a game in 6502 assembly

>>56128743
Is it pong?

>>56128708
Safe pointers mainly

>>56128736
Just use premultiplied alpha you tard

>>56128736
w-what?
also averaging 4 pixels literary results in the highest possible mipmap quality

>>56128490

If I did this in C instead of using cython, could I compile it as an executable and use it?

File: no.jpg (186KB, 816x488px) Image search: [Google]

186KB, 816x488px

>>56128616
>Null terminated strings

I think part of the reason people likes C# is not because of the language itself, but VS. am I wrong? have you tried writing C# code in, say, vim or even notepad?

File: sicploli.png (631KB, 1111x1600px) Image search: [Google]

631KB, 1111x1600px

>>56128877
http://nuitka.net/pages/overview.html

File: Only_the_dead_can_know_peace_from_this_evil.jpg (41KB, 500x490px) Image search: [Google]

41KB, 500x490px

>ask question on stackexchange.com
>immediately gets downvoted to oblivion
>"dupe"
>link to another question that doesn't help
>question gets closed

>>56128883
I use C# because it's nice and clean compared to C/C++
I don't use Visual Studio

>>56128937
>things that did not happen

>>56128827
I can't.

>>56128836
I can average the RGB parts just fine, but I can't average the A parts, because then the fragment shader gets wrong data. The custom alpha channel stuff I use does not directly describe the pixels' alpha level. If I take four pixels and blend them together, the final generated value will be invalid.

This all is part of a highly experimental order independent transparency system I created. It works as long as you don't mipmap.

>>56128883
I'm not sure that I would write C++, Python, or Java outside of an IDE either.

I regularly use things other than VS to write C#, with a commonality in that they all have IntelliSense.

I don't like writing code of any language in something so bare-bones as notepad.

Also, Visual Studio supports most major languages, so the point in using C# just so you can use VS is completely moot.

>>56128937
>ask question on stackexchange.com

I've found the problem

>>56128883
Same can be said about java. But C# > Java

>>56128937
Post your question.

>>56128960
>can't use premultiplied alpha

Enjoy your shit quality I guess.

>>56128879
exactly

>>56128879
Text should just be treated as blobs.

>>56128883
I use C# because it's much more comfy than java. And faster, at least on windows. Visual Studio is nice, but I can't wait until JetBrains relase Project Rider

>>56128902

yeah but let's say that I know I could this in C in a couple hours and want the performance of pure C

>>56129035
IntelliJ is amazing. I'll definitely start using Rider.

So here's a topic for discussion, I have been messing around with C# and flipping through the book Pro ASP.NET MVC 5. There's a big section on dependency injection and then using Ninject. I get the reasoning behind DI but isn't this all going a little too far? Implementing all of these systems and subsystems just to reduce coupling (while adding a ton more complexity)?

>>56128736
>>56128960
I think I kinda understand what you're trying to do now
To my knowledge mipmaping is all automatic and you can't really influence it though
So you probably must use another texture

>>56128879

Pascal did strings right with length prefix, before C ruined everything a few short years later.

>>56129087
>isn't this all going a little too far?
It is.
Next topic.

>>56129087
Like most of the other things you see people complaining about on Java/C#/etc., these are things that are only helpful in massive codebases, that go through massive changes after production, built by massive teams of hundreds and thousands of developers.

Most developers will never need to personally bother with dependency injection.

>>56129087
What you see is consequence of OOP and a reason why it should be purged from face of the earth. And all those who belive in it burned at stakes.

>>56129139
>1 byte string length
>STRINGS LONGER THAN 255 ARE NOT ALLOWED

>>56129087
You still have coupling with dependency injection. It just looks prettier on the surface.

>>56129185

Let me save this thread 50 posts:

>OOP is always horrible!
>OOP is always good!
>OOP can be horrible if you use too much of it!
>OOP can be good if you use the right parts of it!
>OOP is defined by the following assumptions!
>OOP is nothing like what you described!
>OOP is shit!
>Pajeet!
>NEET!

>Something something Haskell/C/Lisp!

>>56129220
kys

>>56129209

This is how I felt as I was reading through the chapter. It seems like you're taking a relatively simple case of coupling and then adding another 30 places for it to fail which are more complex to understand than the original slightly more coupled solution. I could see injecting dependencies but getting to the point where you're using Ninject seems absolutely asinine to me.

I'm a relatively junior developer though so I figured all these senior developers etc. who have spent all this time and effort making these systems must have done so for a reason.

File: makiii.jpg (556KB, 1291x1381px) Image search: [Google]

556KB, 1291x1381px

>>56129220
Well put, have Maki

>>56129220
top kek, 1/4 of dpt resumed in a single post. congrats!

fun main(args : Array<String>) {
  val scope = "world"
  println("Hello, $scope!")
}

>val
Because we didn't already have auto, let, var, ect...
Why does every new language needs it own meme way of declaring variables?

>>56129272
val stands for value while var for variable its completely sound identifier.

>>56129272
What lang?

>>56129282
Doesn't a variable hold a value?

What's a value?

>>56129282
Both put a value in an identifier. You don't need to keep using new keyword for it.

>>56129287
Kotlin

>>56129238
We (the Android team and I) use Dagger for dependency injection in Android apps. The main reason is to make singletons replaceable for testing purposes.
Our network layer is an interface. We have two classes that implement the interface. One connects to the endpoint and the other returns mock data. The latter is used during testing for consistency.
Android doesn't give you the option to pass the instance of the network layer due to the horribly bad designed SDK. So you're stuck with using singletons or dependency injection.

>>56129201
2 byte string length then, not that hard

also
> STRINGS CANNOT CONTAIN A ZERO BYTE

>>56129303
>>56129310
val x = 3;
val binds value 3 to identifier x (unmutable)
x = 4 => error
var x = 3
x = 4 =>fine

>>56129359
How about variable length word machines?

>>56129368
Why not just call it a constant?

>>56129374
i don't see how that's relevant, just because your words have to be terminated in some way doesn't mean a string has to be null terminated.

>>56124255
That framedrops...

>>56129397
you limit the possible length of a string.

>>56129368
that's what let is for

>>56129388
Dunno its language design choice, guess its shorter. But as name for new identifier its completely fine.

>>56129411
who limits it?

>>56129428
You, scambag

>>56129416
shorter than "let"?

>>56129415
Scala doesnt have let, also let expression in lisp isnt used for constants but for defining binds within expression

>>56129434
shorter than "constant"

>>56129433
by not having variable length words?
ok whatever
it's not that unreasonable to just have a 32-bit int string length and you'd almost never exceed that length.

>>56129457
No, you are retarded and you can't even read properly.

>>56129468
please explain your point then, because i'm pretty sure you haven't.

>>56129480
>because i'm pretty sure you haven't
I haven't what? You are not making any sense.

>please explain your point then
Systems with variable length words do not enforce limits on the size of memory, while you are forcing arbitrary limits to the string size.

>needing more than 255 characters
justr use a stringing

File: YeOzLpv.png (338KB, 732x912px) Image search: [Google]

338KB, 732x912px

is it worth it /g/ ?

>>56129517
Like you'll read all of them.

>>56129508
> hypothetical turing machines do not enforce limits on the size of memory, while hardware manufacturers are forcing arbitrary limits to memory size

just use a 32-bit string length you dumbass

>>56129517
I probably wouldn't do anything past the $1

>>56129517
>If Hemingway Wrote Javascript

Jesus christ, how horrifying.

>>56129541
Maybe 8 for lisp books, so he can reach enlightenment

>>56129517
>>56129534
>>56129541
Ironically the $1 all seem good and the rest all seem trash

>>56129535
>hypothetical turing machines
Real machines, faggot.

>just use a 32-bit string length
"32-bits should be enough for everyone!" - you

>>56129553
In fact I'd make sure to only pay $1 so that you don't accidentally read any of the shit by mistake

>>56129555
32 bits is more than enough especially if you are dealing with unicode characters.

>>56129555
if you need a long string you can use 64-bit if you really want.

but if you need a longer string you should likely be using a different data structure anyway

>>56129555
>I need a 2^32 length string
pls

>>56129555
>32 bit string isn't enough
>4.29 billion characters

>>56129583
4.29 billion 1 BIT characters which changes if you dont use ASCII

>>56129583
>4.29 billion characters
You might want to try that math again, friend, and specify the encoding.

>>56129597
1 byte, not 1 bit

>>56129609
I was assuming 1 byte per char

>>56129597
>>56129609
Even with Unicode, that's still over a billion characters. When is it the case you'll ever be hitting that many in a string?

File: 1460235856510.png (158KB, 379x440px) Image search: [Google]

158KB, 379x440px

Is Learn Python the Hard Way a good place to start?

I dabbled a little bit in C a while back so I'm not completely clueless, but I'm just wondering if its good for beginners.

>>56129644
its only 536870912 chars with unicode

>>56129622
>>56129640
A byte has variable size so you do not make sense.

>>56129650
It is but you should stick to C

>>56129650
Zed Shaw is a hack.
His teaching style is effectively
>copy this large program
>now change some variables until it breaks
>explain why the program broke
>next chapter!

>>56129665
>le C language meme

>>56129653
2^32 / 4? UTF-8 uses up to 4 bytes.

>>56129661
my dick has variable size

>>56129661
>a byte has variable size
???

>>56129678
No its not about C, its about sticking to one lang when you are noob, mixing languages is from pedagogical pov.

>>56129681
i took largest size, with 4B its 1073741824

>>56129696
mixing is bad*****

>>56129692
Exactly.

>>56129650
Learn Python the Hard Way is fucking garbage because Zed places all of the work in your lap instead of a student/teacher division of teaching and working. Instead of explaining concepts, demoing them while giving practice snippets, and then asking for recreation, he goes: "See all these print statements? Type them out into your editor. It jus werkz. Now change this. Now it doesn't werk. Why? :^)"

Automate the Boring Stuff is a superior book. It covers what you would expect but doesn't get deep into the language. However, it allows the user to meddle with basic concepts and be told information instead of being told "type this collection of print statement garbage I already made out it'll make you SMERT".

>>56129692
anon >>56129728 has a point
>The size of the byte has historically been hardware dependent and no definitive standards existed that mandated the size.

>>56129728
>Python

They're both shit

>>56129740
ffs not >>56129728
>>56129716 this one

>>56129740
That's machine specific not variable size
The size doesn't change at runtime

And anon was complaining about 32 bit index strings not being long enough and wanting 64 bit index strings

The overwhelming majority of 64 bit platforms are i86-64

hey

guess what?

NEW THREAD!

>>56129785
>>56129785

File: cover.jpg (51KB, 400x579px) Image search: [Google]

51KB, 400x579px

why do people think this book is a meme?

> I don't want to learn
> - how to write an interpreter
> - how to write a compiler
> - how to write a logic programming language
> - how non-determinism works
> - how lazy evaluation works
> - how to write a garbage collector

How do you come up with a n logn algorithm

>>56129789
because it forces you to write these things in a dead language from the 1970s

File: 1461355444750.png (48KB, 1024x874px) Image search: [Google]

48KB, 1024x874px

>>56128937

>>56129800
Yet it happens to have more features while staying bloatless in comparison to many "modern" languages.

>>56129742
Thanks for sharing.

>>56130030
You're welcome

Remember: don't use Python

>>56130035
You're not the boss of me.

>>56129800
>>56130018
Also, why do people whine so much about languages? You could easily use sicp with your own or whatever.

>>56130050
How?

>>56130102
By not being a retard, not something YOU can do.

File: shiny mad bunny.png (252KB, 660x700px) Image search: [Google]

252KB, 660x700px

>>56123479
>current year berserk