/dpt/ - Daily Programming Thread

Probability of Python 3 becoming an ISO standard?
>many supported platforms
>widely used and deployed
>multiple major implementations (CPython, IronPython, Jython and PyPy)
>multiple cpython variants
>multiple sub-set implementations (Brython, CLPython, HotPy, pyjs, PyMite, pyvm, RapydScript, SNAPy, tinypy)
>multiple implementations in progress (Berp, phpython, Pyjaco, Pystacho, pyvm2, Skulpt, Typhon)
>shitload of libraries and frameworks
>many working products made

File: 1496810768852.jpg (73KB, 398x398px) Image search: [Google]

73KB, 398x398px

>>62162915
Please post a _cute_ anime image next time.

File: 1472937160601.jpg (45KB, 640x480px) Image search: [Google]

45KB, 640x480px

>>62162975
if you don't love lain then you shouldn't be on /g/

>>62162967
What's the benefits tho?

>>62162983
Lain is fucking garbage. Piss off with your forced meme.

File: 1479179758186.jpg (35KB, 519x517px) Image search: [Google]

35KB, 519x517px

>>62162991

File: Trials_Fusion_Editor.png (1MB, 1278x719px) Image search: [Google]

1MB, 1278x719px

Why don't PCs have the ability to fully render code in 3D Space?

Trials Fusion absolutely nailed this, and taught me more about the basics of programming than any text could have.

>>62162991
>t. trapposter

>>62163002
Why not 4D?

>>62162985
Languages without a respectable authority standard will die in the long term.

>>62163002
Because it's inconvenient to program on a fucking 2D monitor in a 3D space. Well done writing that ten line program in the screenshot.

>>62163018
And by respectable authority I mean ISO/ANSI/IEC/POSIX.

>>62163002
Non-textual representations of code fall apart and become stupidly hard to navigate once you get past trivial programs.

>>62163003
How does thinking lain is not cute imply that I'm some of those filthy degenerates?
I normally post the Yuki flipping image.

                                                                                                                                        I've never even bothered to watch lain. I just call it shit to trigger you fags.

>>62163018
Everything will die in the long term.

>>62163042
No anon, you're talking about the very long term, not about the long term.

>>62163030
https://www.iso.org/ics/35.060/x/
Some languages died already, yet they had ISO standard

>>62163018
"Implementation is the standard" is better than design by committee IMO.
>will die in the long term
No? Java, Python, PHP, Perl, Ruby and VB aren't standardized and they all are alive languages from TIOBE's top 10 with long histories behind them.

>>62163062
Well, they can be resurrected.

>>62163067
They would be safer with a commitee standard though.

>>62163018
cont.
As a counter example, language like Forth, Common Lisp, Smalltalk and Pascal has been standardized, yet are virtually dead nowadays.

>>62163058
Long term is the very long term.

>>62163079
>>They would be safer with a commitee standard though.
Someone get this hothead outta here.

>>62162915

#include <stdio.h>
#include <math.h>

const double table[] = {
  .00, .01, .01, .02, .02, .03,
  .03, .03, .04, .04, .04, .05,
  .06, .06, .07, .08, .09, .10,
  .11, .13, .14, .16, .18, .20,
  .23, .25, .29, .32, .36, .40,
  .45, .51, .57, .64, .72, .81,
  .91, 1.02, 1.14, 1.28, 1.44, 1.61,
  1.81, 2.03, 2.28, 2.56, 2.87, 3.23,
  3.62, 4.06, 4.56, 5.12, 5.75, 6.45,
  7.24, 8.13, 9.12, 10.24, 11.49, 12.90,
  14.48, 16.25, 18.25, 20.48 };
const unsigned tablen = sizeof(table) / sizeof(table[0]);

double expt(double t, double x) {
  return (pow(t, x) - 1) / (t - 1);
}

int main() {
  double min = 500;
  double maj = 5000;
  double a = (maj + min) / 2;

  double y = table[tablen - 1];
  double n = tablen - 1;

  unsigned itermax = 1000;
  for (unsigned iter = 1; iter <= itermax; ++iter) {
    double d = 0;
    for (unsigned i = 0; i < tablen; ++i)
      d += y * expt(a, i / n) - table[i];
    fprintf(stderr, "(%u) d=%e\n", iter, d);
    if (d == 0) {
      break;
    } else if (d < 0) {
      double old = maj;
      maj = (maj + a) / 2;
      fprintf(stderr, "MAJ %f -> %f\n", old, maj);
    } else {
      double old = min;
      min = (min + a) / 2;
      fprintf(stderr, "MIN %f -> %f\n", old, min);
    }
    a = (maj + min) / 2;
  }

  fprintf(stderr, "A=%.17e\n", a);
  for (unsigned i = 0; i < tablen; ++i)
    printf("%u %f %f\n", i, table[i], y * expt(a, i / n));

  return 0;
}

>>62163079
> They would be safer
What does this even mean? A committee can't force anyone to use the language it maintains, and the success record of the committee-maintained languages is mixed at best.

File: trialsfusion1.jpg (76KB, 1280x720px) Image search: [Google]

76KB, 1280x720px

>>62163024
>>62163034
I disagree. Being able to see everything on the screen and how they're connected works a treat. You even get to see visual representations of the impulses being triggered.

https://www.youtube.com/watch?v=i93d1QfTOK4

https://www.youtube.com/watch?v=qz_8C1zqxn4

https://www.youtube.com/watch?v=_IJvFHvc-h8

https://www.youtube.com/watch?v=5Par5zpjiFQ

https://www.youtube.com/watch?v=DFUaWe56eYw

https://www.youtube.com/watch?v=6jyMX3-au5c

Even if it's not your personal cup of tea, you can't deny it's an excellent way of visually showing how programming works, at very least on a basic level.

>>62163120
> const unsigned tablen = sizeof(table) / sizeof(table[0]);
The type of `sizeof` is `size_t`, not `unsigned`.

>>62162915
Can we all agree Lain programs in C?

>>62163159
No, because she programs in Lisp.

>>62163147
If you're a baby who needs visual effects to keep his attention when learning programming, maybe.
For productive work, this is worthless.

File: lain_lisp.jpg (129KB, 1366x768px) Image search: [Google]

129KB, 1366x768px

>>62163159
>he hasn't watched the show

why do we even need signed ints
is there ever a time where you cant just change your function slightly and do without them
there isnt even such thing as a negative number

>>62163174
Common Lisp is the hackiest language

>>62163191
Balance can't be negative? Nice!

ISO is the reason why C++ is worse than it could've been.

>>62163168
What is it about the word 'learning' you're too stupid to understand?

>>62163209
ISO is the reason why we have C++ and not C++/MS, C++/GNU, C++/Intel, C++/Apple, etc.

>>62163147
That is still nowhere near the level of "real" software and programming.
Scale that up to the equivalent of 50,000 lines of code (what would be considered a "medium" sized program), and tell me how it works then.

>>62163206
Yes it can't, it's called overdraft, not negative balance.

>>62163234
>C++/MS, C++/GNU, C++/Intel, C++/Apple, etc.
You DO have that, though. In fact, it's far worse than that.
There are literally thousands of different versions of C++.

>>62163221
I did use the word learning in correct context in my reply. Baby.

>>62163259
rewrite that in a decent language with good libraries and it'd be 10k

>>62163147
How the fuck am I supposed to make a kernel with this trash?

>>62163269
Wow, that is a really fucking stupid statement.
There is literally no context for what I said about the 50,000 lines of code.

>>62163259
For BASIC programming.

>>62163262
>6.
> a figure representing the difference between credits and debits in an account; the amount of money held in an account.
When credits are greater than debits, balance is negative.

>>62163266
>You DO have that, though. In fact, it's far worse than that.
No, it's not? Practically any major C++ project, like Qt or Chrome, can be compiled with any standard-compliant C++ compiler with minor to no changes.
> literally thousands of different versions of C++
What?

>>62163291
so you're saying its the difference between two numbers
.... which is a positive int

>>62163234
We have that. They are incompatible between each other. Both at source code level and at binary level, even for same platform.

>>62163302
You're saying that owing 1000 dollars and being owed 1000 dollars results in the same balance value - one thousand.

>>62163312
yes the value of money is 1000 for each

debt is a capitalist structure

>>62163316
That is incorrect.

>>62163301
>What?
Literally every psychopath (i.e. sepplesfag) has their own "C++", with whatever different shit they've picked and excluded.
You can take the ""C++"" written by one """C++""" programmer, and how it to a different """"C++"""" programmer, and they will be completely unfamiliar with it.
That's why there are literally thousands of different versions of C++.

>>62163305
>They are incompatible between each other.
They all are compatible with each other as far as the standard goes, ABI incompatibility is not a language issue.

>>62163324
1. '-' isn't a digit so you cant use it as part of a number
2, you cant have negative distance to somethinng dumb dumb
3. balanceis caculatted as the distance between two numbers, which is a positive value

negative values do not exist

File: Icon-Sat-V2.png (469KB, 790x720px) Image search: [Google]

469KB, 790x720px

>>62163147
Fuck even is this?
These videos are boring as hell.

>>62163167
>>62163174
Yup, definitely Lisp.

Kinda want to learn that, now, desu.


BTW, if anyone is interested in a /g/ Server on Discord, we got one over here:
https://discord.gg/csEwM

Global type of /g/ for programming, machines, game creation, Science and Maths, etc. And, ofc, chatbot programming

>>62163301
>with minor changes
Well done admitting there actually are different incompatible vendor-specific versions of C++.

>>62163338
>They all are compatible with each other as far as the standard goes
And this is irrelevant, because people do use non-standard features.

>>62163343
You're confusing distance and difference. You're actually just a very confused person. This is my last reply to you.

>>62163361
im right
dont reply

>>62163327
Committee denies it. The embedded community has been asking for the standard to give us a method for constraining the feature set for years now.
I doubt the committee would like to add something like this to the spec (despite them bloating it with abandon) just because it'd make what you describe plainly obvious. It'd be problematic from a PR perspective. Right now people blame each other for writing 'wrong C++'.
If you let the language exclude features or styles you'd shift the blame to the language because it's so unwieldy.

>>62163079
lol, no. Commitees are actually the worst. They restrict the development of a language because any new features in a language require that a committee come to delegate and decide new features. Then each member of the committee want to put their own favorite features into the language and you get a cluster fuck up of a language that is basically inconsistent. See C++ and how messy the language is and how long it takes to change any shit in the language. Especially when you have corporations on the committee who don't want to break backwards compatibility with their product and will push to have dumb shit still in the language. Languages are much better designed by a single person or small group of people because then the language has direction and can remain consistent and not have dumb shit tacked on. Though its too late for Python anyway because Python is already a hacked together language which I don't particularly like.

>>62163347
>proprietary software
Trash
Call me when you'll have FOSS solution.

>>62163327
This is a meaningless remark, we are talking about implementations, not about styles, it's still the same language whatever you use spaces or tabs.
>>62163350
>And this is irrelevant, because people do use non-standard features.
Yet every major C++ project can be built with any modern C++ compiler tho, you can use any C++ book with any modern C++ compiler, most of the libraries work with any modern compiler. For some reason, you keep ignoring the reality out there.

>>62163395
Well, it's free as in free beer.
And there are plugins, so it's kinda free as in Free Speech

>>62162915

#include <stdio.h>

int main(){

int c = 5, d;
d = ++c + ++c + ++c + ++c + ++c + ++c + ++c + ++c + ++c + ++c + ++c + ++c;

printf("%d\n", d);

return 0;

}

How much is d?

>>62163396
>Yet every major C++ project can be built with any modern C++ compiler tho
Yes, and they have people dedicated to ensuring that their projects can be built with those compilers. Your point?

>>62163409
UB lol

>>62163409
d is free

>>62163415
> Your point?
Ok, whatever, this is going nowhere.

>>62163409
It's >= c.

(defvar d
  (let* ((c 5))
    (+ (incf c) (incf c) (incf c) (incf c) (incf c) (incf c)
       (incf c) (incf c) (incf c) (incf c) (incf c) (incf c))))

(format t "~A~%" d)

How much is d?

>>62163448
False

>>62163234
The only C++ that should exist is C++/Stroustrup
Seriously, C++ would be so much better if the committee bothered to actually fucking listen to him.

File: 1503937708014.jpg (1MB, 500x2855px) Image search: [Google]

1MB, 500x2855px

Read the right books, anon.

>>62163475
This is a really shitty maymay.

>>62163536
t. underage.

>>62163385
>They restrict the development of a language
>Though its too late for Python anyway because Python is already a hacked together language
Wouldn't this protect Python from more shit added, or at least slow that process down?

      PROGRAM EUCLID
        PRINT *, 'A?'
        READ *, NA
        IF (NA.LE.0) THEN
          PRINT *, 'A must be a positive integer.'
          STOP
        END IF
        PRINT *, 'B?'
        READ *, NB
        IF (NB.LE.0) THEN
          PRINT *, 'B must be a positive integer.'
          STOP
        END IF
        PRINT *, 'The GCD of', NA, ' and', NB, ' is', NGCD(NA, NB), '.'
        STOP
      END

      FUNCTION NGCD(NA, NB)
        IA = NA
        IB = NB
    1   IF (IB.NE.0) THEN
          ITEMP = IA
          IA = IB
          IB = MOD(ITEMP, IB)
          GOTO 1
        END IF
        NGCD = IA
        RETURN
      END

>>62163503
Really? How so, I don't exactly follow C++ internal politics.

How bad is it that my html parser or whatever the term is, is O(kn) where k > 1 ?? I have to iterate through the damn webpage multiple times.

>>62163582
No, what do you mean? You can't fix a fucked up language with a committee, a committee can only make it worse. I mean, Python 3 is fucked because people still live in Python 2 land and standardizing the language isn't going to make them switch. Also, not having a committee means you can change the language much easier because you just change the implementation, see Ruby. Ruby implements changes to the language, but they don't split because nobody bothers supporting the old shit, which they would if the language was standardized. Standardizing Python which just create a further split in an already fractured community. Also, Python is on its way out in popularity outside of the scientific community and is getting replaced by Go in a lot of cases.

>>62163424
>>62163427
>>62163448
Incorrect. You suck balls.

>>62163649
>and is getting replaced by Go in a lot of cases.
I don't know whether to cheer or cry.

>c++17
>most codecvt stuff got deprecated
What the hell replaces std::wbuffer_convert?

>>62163301
Yeah, just like C can be "portable". For some reason you will get fuck wits who program outside a standard and use not standard features for convenience.

>>62163473
What kind of moonlanguage is that? Learn a real language, faggot.

>>62163649
>No, what do you mean? You can't fix a fucked up language with a committee
You can slow down the language getting even more fucked up.

Also, a language that breaks backward compatibility with the predecessor, however small that change would be, is a new language.

>>62163673
>doesnt know lisp
You're a fucking idiot.

>>62163688
That's not a language. That's a family of languages.

>>62163692
the l stands for language

>>62163475

You got it backwards I think..?

BTW totoro was one of the most disturbing movies ever..

>>62163670
> For some reason you will get fuck wits who program outside a standard and use not standard features for convenience.
This is the reason the kernel can only be compiled with gcc. Say that you want about Linus, but not insisting on compiler portability from the start was a huge mistake.

GUYS STOP ARGUING ABOUT WHICH LANGUAGE IS THE BEST AND ANSWER MY FUCKING QUESTION

ANY LANGUAGE IS BETTER THAN ANY OTHER LANGUAGE DEPENDING ON THE FUCKING USER SHUT THE FUCK UP

>>62163700
no, l stands for list in list processing

>>62163602
I only know he's been trying to add Concepts to the standard since ~2009.

File: pineal-gland.jpg (29KB, 600x408px) Image search: [Google]

29KB, 600x408px

Are there any intelligent people here?

>>62163721
I can live without Concepts. I can't live without modules or UFCS.

>>62163710
no its not and your a retard for believing that

>>62163730
no
now fuck off

File: akarireaction.jpg (28KB, 344x344px) Image search: [Google]

28KB, 344x344px

Clueless idiot who's never done this before reporting. I've set up a Google Sheets file that nabs the value of a particular currency using the following:

=IMPORTXML("https://coinmarketcap.com/assets/COIN-NAME/","//*[@id='quote_price']")

It works just fine, but Google Sheets only refreshes the value after a long time, something like 20 or 60 minutes iirc. So I'm trying to figure out how to make a script that forces a refresh every 5 minutes or so, and what I've come up with is:

function getData() {
  var queryString = Math.random();
  var cellFunction = '=IMPORTXML("https://coinmarketcap.com/assets/COIN-NAME/","//*[@id='quote_price']" + queryString + '",1);

  SpreadsheetApp.getActiveSheet().getRange('Logs!A1').setValue(cellFunction);
}

But of course I don't know how any of this works and just keep getting the error: Missing ; before statement. (line 3, file "Code").
What's the correct way to get this working? Although the following works, var cellFunction = '=IMPORTHTML("<url>?' + queryString + '","table",<index>)'; it brings up a whole table of data I don't need, all I want is a single cell that would bring out the exact result that the first function of this post brings.

>>62163718

B T F O
T
F
O

File: t2.png (214KB, 355x453px) Image search: [Google]

214KB, 355x453px

>>62163738
Why? Do you not seek the path to enlightenment?

How many languages are you an expert in?

>>62163666
Cheer. Go is a more well thought out language by some of the brilliant minds at Bell Labs with Rob Pike and Ken Thompson and wasn't a holiday hobby project by some amateur language designer.

>>62163678
>You can slow down the language getting even more fucked up.
Doesn't really matter in the end, a fucked up language is a fucked up language and slowing it down isn't going to change shit.

>a language that breaks backward compatibility with the predecessor, however small that change would be, is a new language.
This is being too technical and unnecessary. For all intents and purposes, the language is the same across versions unless it breaks non trivial backwards compatibility.

>>62163735
> UFCS
I know that it is but I can't see why people care about it, it's just minor sugar, besides we already have std::mem_fn.

File: 1503882284711.jpg (178KB, 634x640px) Image search: [Google]

178KB, 634x640px

Did you enjoy Ruby on Rails?

>>62163752
>This is being too technical and unnecessary. For all intents and purposes, the language is the same across versions unless it breaks non trivial backwards compatibility.

for all intents and purposes C++ and Python are good languages because you can use a good subset of both. your argument is a slipper slope

>>62163752
Static typing with no generics is just too painful. Handling void*s is my least favourite part of C.

>>62163766
>for all intents and purposes C++ and Python are good languages because you can use a good subset of both.
I fully agree, unironically

>>62162915
why is this thread callee
>/dpt/ - Daily Programming Thread
and not
>/Programming General/ - Programming General
?

>>62163520
>Fortran is somehow higher than C++ or F#
You haven't programmed in Fortran in your lofe, haven't you?

>>62163766
I can use a subset of C++ and it would basically be C, it doesn't mean that C++ is C. I could use an Objective-C compiler to write a C programming language, doesn't mean that Objective-C is C because it isn't idiomatic Objective-C.

Wew, they've just added experimental generators to nightly Rust: https://doc.rust-lang.org/nightly/unstable-book/language-features/generators.html :

#![feature(generators, generator_trait)]

use std::ops::Generator;

fn main() {
    let mut generator = || {
        println!("2");
        yield;
        println!("4");
    };

    println!("1");
    generator.resume();
    println!("3");
    generator.resume();
    println!("5");
}

I hope it actually makes programming with Futures/Tokio bearable.

>>62163777
Because it has this name from prehistoric times.

>>62163806
You haven't worked in projects where each programmer chooses his own "good part" or the language.

>>62163775
What point are you trying to make?

>>62163826
Fuck, mixed those up.

>>62163826
I don't know what you mean by
>your argument is a slipper slope

>>62163826
>What point are you trying to make?
That I actually hold that opinion

Anyone able to point me in the direction of a solution to the knapsack problem where there is a limit on both weight and value.

Say I have limit of $100 and two types of items, one weights 20 and gives 10 profit, one weighs 5 and gives 1 profit.

If I put a limit on only being able to use 4 of each item I get, 20+20+20+20+5+5+5+5.

I already have a working unbounded problem which would just go 20+20+20+20 as this is the highest value you can obtain, but I have no clue how to bound the knapsack while not doing 0/1 bounding

>>62163826
The point I'm trying to make that a language is basically the same across versions. I can write a program using Go 1.0 but I can still compile it on a Go 1.9 compiler. They are basically the same language and it is unnecessary to say they are different because Go 1.9 doesn't break any backwards compatibility with Go 1.0.

File: 15006286503.jpg (163KB, 1024x768px) Image search: [Google]

163KB, 1024x768px

PROGRAMMING CHALLENGE

Make a £sd calculator. It must be able to add, subtract in £sd, multiply and divide by decimal fractions. Lowest coin is farthing.
4 farthings make 1 penny, 12 pence make 1 shilling, 20 shillings make 1 pound.

So far, we have a solution in APL, J, C and C++!
What other languages will present /g/ later?

>>62163838
oh,
you're right, I haven't worked in projects with C++ or Python where several people choose their own "good parts"
but that's irrelevant because I'm a freelance developer and I build all of my clients software myself
so, to me, they are both excellent languages with all of the good parts included and all of the bad parts essentially don't exist

>>62163854
>one weights 20 and gives 10 profit, one weighs 5 and gives 1 profit.
what

>>62163803
I have, and that's exactly how I know.
By all means, continue being the low-IQ software developing pleb you are. Leave proper research to those with a grain of intelligence.

>>62163440
It stayed where it begun. There are plenty of vendor-specific C++ implementations, each providing different incompatible between each other feature-set, and people do use those features. You mistakenly thought that the fact that some major libraries can be compiled on any of popular compilers meant that ensuring this situation is possible is free. It's not. Your assumption about committee being helpful is incorrect also.

>Start writing an interpreter for a VM
>Do it in Rust because why not
>End up having to write asm anyway
I should have used C.

>>62163911
>Leave proper research to those with a grain of intelligence.
So, Python is the language of übermenschen?

File: g's Official Language Chart.jpg (1MB, 620x4144px) Image search: [Google]

1MB, 620x4144px

Here are the official programming intelligence requirements of /g/.

>>62163942
>/g/'s Official Language Chart is a facebook meme
Fascinating.

>>62163092
Pascal dead? How about fuck you

>>62163962
Dead and buried. Rip.

>>62163921
Falling for memes, huh?

>>62163998
Actually it's quite good.

>>62163962
Why so aggressive, Niklaus? It's been dead for decades now.
>>62163921
Why do you need asm for this, are you doing your own JIT?

>>62163895
20 and gives 10 value, one weights 5 and gives 1 value.

Sorry, i'm using price and profit instead of weight and value respectively.

>>62164027
No, I'm writing a Piet interpreter than can do syscalls. I couldn't find a way to pass a variable number of arguments to the syscall! macro most crates propose so I have to manually put my shit in the right registers and call the interupt.

I finally figured out a n^2logn solution to a problem in my algs&data structures class and I'm so fucking happy.
The problem is finding the maximal collinear line segments in a set of random points in n^2logn worst case with space proportional to n + number of line segments.

Supposedly it's doable in n^2 but it's a very complex algorithm that involves collecting every line segment and doing a topological sweep. No clue how that black magic works

>>62163942
Worst language chart ever. kys

How do I run c++ programs on Windows? I use sublime text since don't want to install visual studio and no other ide seems to work for some reason.
I downloaded some cygwin thingy still it shows some error.

>>62164046
rust doesn't have inline asm?

>>62164046
> I couldn't find a way to pass a variable number of arguments to the syscall! macro
https://doc.redox-os.org/kernel/src/x86/syscall.rs.html#16-40 ? Granted it still uses asm to do syscalls because what else, but at least you can use nice syscall!(whatever, a, b, c) macro.

>>62164082
install visual studio
if you're not going to use visual studio, use another language

>>62164082
>it shows some error.
Try the solution.

>>62164040
You have to get the most weight for $100?

>>62164085
It does but having to write asm is what I'm complaining about.

>>62164093
I don't always have the same amount of arguments. Sometimes I have 1, sometimes I have 2. Sure, I could just do a match on the len of my arguments vector and call the macro in each branch but by then I might just as well write my own asm.

>>62164096
I can download nine different Linux distros for that file size.

>>62164113
2 or more

>>62164113
wait wait
rust doesn't have function overloading?

>>62164121
deal with it

>>62164132
It has generics and variadic functions.
What I would have liked is a method on my vector that takes a function as argument and automatically calls that function with each element of the vector as argument.

>>62164113
It makes sense for an interpreter since you don't know the number of the arguments at the compile-time.
>>62164154
>What I would have liked is a method on my vector that takes a function as argument and automatically calls that function with each element of the vector as argument.
You mean `map`?

>>62164185
No, map applies the function to each element. I want the function to be called only once, with every element passed as argument. Sorry if I'm unclear, english isn't my mother tongue.

How do you feel about enums?

>>62164216
They're great if they aren't just treated as numbers by your compiler.

>>62164111
The price can be set to anything the user desires, and the amount of times an item can be used is also set by the user.

Price is the same as weight.
Profit is the same as value.
(in the traditional knapsack problem)

So I have to write a dynamic programming algorithm to correctly pick the items that make up the highest profit while not using them more than i'm allowed. So an example is:

Item 1 = $20price, $10 profit.
Item 2 = $5price, $1 profit.

amount of times I can use an item: 4
price limit to be used: $100

so, $20 + $20 + $20 + $20 + $5 + $5 + $5 + $5 is equal to $100 and gives profit of $10*4+$1*4 or $44 profit.

I have it working for unbounded problems where the item can be used an unlimited amount of times, but I have no idea how i'd go about limiting by X amount of usages.

>>62164200
I see, IMO the best way is to pass the vector as an argument and then match it to call syscallN functions, something like

fn syscall(syscall_no: u64, args: &Vec<u64>) -> u64
match *args {
  [] => syscall0(syscal_no),
  [a] => syscall1(syscall_no, a),
  [a, b] => syscall2(syscall_no, a, b),
  ...
  _ => unreachable!(),
}

>>62164240
>but I have no idea how i'd go about limiting by X amount of usages.
ok I'll give you an idea
sort the items from best profit to worst profit
then take as many of the best items as you can
then take as many of each of the remaining items, in order, until you have reached the money limit

>>62164252
Yeah, that's what I was taling about here >>62164113 but then why even bother depend on a crate to do that. The interpreter isn't portable anyway.

>>62163688
Oh, just another SICP faggot. Go light youself and your stupid book on fire.

>>62163710
Best languages ever are C, Java and Python. End of story. That's why those languages are widely used and have been active for years. Languages such as Go, Ruby, F# or any other shit come and go.

>>62164259
So just a greedy approach? I have to do it using dynamic programming so I'm not entirely sure if I'm allowed to do it that way :/

My current idea is to have a count checking how many times i'm using an item, once it's been used too many times i'll remove the item from the list, hopefully that won't break my memoization list

>>62163921
> using Rust

>>62163942
> Lisp, the 4th most challenging

My fucking sides.

>>62163710
>EVERYONE IS BEAUTIFUL
>EVERYONE IS SPECIAL
>EVERYONE GETS PARTICIPATION TROPHY
>WE NEED MORE DIVERSITY
>REALITY IS OPPRESSIVE
Wew.

>>62164290
yeah, greedy algorithm is usually the solution used for knapsack problems
unless you need the best solution

I can't really think of a reason to use dynamic programming for this
but since you're required, I'm betting there's a clever dynamic programming knapsack algo that you're supposed to find/figure out and implement

File: phonebook.jpg (83KB, 701x381px) Image search: [Google]

83KB, 701x381px

This makes no sense.
Linear search is O(n) and so is the "in" operator.
Since unique_names is as large as phonebook, in the worst-case, the performance of the whole function should be O(n^2).

>>62164520
>python
>anything clever
What would you except from Turing incomplete language?

>>62164287
>Python
>for years

>>62164520
>Linear search is O(n) and so is the "in" operator.
There're no "in" operator here, both `in`'s are part of `for .. in ..` syntax. But you're right, the code is O(n^2).

So i went from desktop dev to android dev, so far so good but it has some annoying as fuck limitations showing it's not a fully matured platform yet.

Time to get 0 downloads on the app store just like all my windows shit got.

Fuck there are way too many programmers in the world, the market is completely oversaturated

>>62163777
yeah, right
/dpt/ was neither daily, nor programming, nor a thread

>>62164579
Shouldn't it be less than O(n) for inner loop since unique names list is initially empty?

>>62164579
You're right, thanks.
I meant the whole loop and not just the operator.

How does decompiling work?

Like for example if someone put a line of code in his application that checks the system date if it's not whatever date it exits the application.
Can i just decompile, delete that one line, recompile and boom i just made a "cracked" app? Does that shit work like that?

>>62164591
>he writes """apps"""
>complains there are too many """programmers""
Mobile dev is like the shareware hype in the late 90s, it's mostly teenagers and shady scammers.

>>62164639
I'd also like to know the answer to this

>>62164520
Yeah it's full of shit. I removed the spurious fluff we don't need.

$ python /tmp/py.py  # 10
1.99791502953

$ python /tmp/py.py # 20
5.53939199448

$ python /tmp/py.py # 40
18.0426440239

$ cat /tmp/py.py 
def list_unique_names(phonebook):
    unique_names = []
    for first_name in phonebook:
        for unique in unique_names:
            if unique == first_name:
                break
        else:
            unique_names.append(first_name)
    return len(unique_names)

pb = [ str(i) for i in range(40) ]

def run_test():
    list_unique_names(pb)

if __name__ == '__main__':
    import timeit
    print(timeit.timeit("run_test()", setup='from __main__ import run_test'))

>>62164579
`in' will call list's membership check which is O(n); that function is literally just

return len(set(map(take_fst_split_name, phonebook)))

which _is_ fast

How can I play with C++17 features today?

>>62164738
https://gcc.gnu.org/projects/cxx-status.html#cxx1z
https://clang.llvm.org/cxx_status.html

>>62164738

-std=c++17

>>62164754
>>62164769
Both clang and g++ won't let me include <variant>.

File: 20_2k.png (940KB, 799x600px) Image search: [Google]

940KB, 799x600px

Back to c++ raytracing

trying to understand what the fuck a surface area heuristic is

>>62164817
Update your compiler, <variant> has been implemented in gcc 7: https://gcc.gnu.org/onlinedocs/libstdc++/manual/status.html#status.iso.201z .

File: 1483889197006.jpg (123KB, 1280x720px) Image search: [Google]

123KB, 1280x720px

>>62164734
Thanks. I checked the book page and it's already listed as a possible error in the "Unconfirmed Errata" section.

http://www.oreilly.com/catalog/errataunconfirmed.csp?isbn=0636920028963

>>62164734
>>62164626
>>62164579
I made some tests and it definitely does seem to be O(n^2), even though I expected it to be somewhat different due to the other list being initially empty.

File: EXCEL_2017-08-30_16-40-18.png (87KB, 1520x625px) Image search: [Google]

87KB, 1520x625px

>>62164944
Forgot the picture. Fug.

I expected it to be <2k but it keeps getting bigger

-------------------------------------------------------------------------------
File                             blank        comment           code
-------------------------------------------------------------------------------
./dg.c                              59              7            628
./tick.c                            54             16            400
./g.c                               12              3            246
./util/share.py                     39             22            208
./g.h                               14             12            197
./dg.h                              22             32            196
./util/p3.py                        32             51            137
./draw.c                            20              0            121
./util/lvlgen.py                    18              8             49
./util/dg.py                        19             15             47
./util/flgen.py                     14              6             40
./main.c                             4              5             22
./Makefile                           5              1             11
-------------------------------------------------------------------------------
SUM:                               312            178           2302
-------------------------------------------------------------------------------

>>62164944
This is why coefficients are ignored in O-notation, it doesn't matter if it's 1/2 * n^2 or 1000 * n^2.

File: 1399077079990.png (255KB, 492x528px) Image search: [Google]

255KB, 492x528px

>>62162915
Anyone here know what the general consensus is on the cs50 edx course? I'm doing the assignments on week 4 but i'm not sure where it's going. I like that it doesn't hold your hand but I feel like it kind of moves at a weird pace, with really simple explanations and notes, then hitting you with assignments that require a lot of other knowledge they only briefly touched on.

pic unrelated

File: EXCEL_2017-08-30_16-44-51.png (91KB, 1524x702px) Image search: [Google]

91KB, 1524x702px

>>62164978
Although!

The plot I posted is for when there can be 3844 different names, so most of names in phone books are unique. I re-ran the simulation with only 62 different names, so there are a lot of repeating ones, and here is the plot. Faster than n log n. It's not as simple as you think it is.

>>62163364
Can I reply?
I too like computer stuff.

>>62164989
>that require a lot of other knowledge they only briefly touched on
typical top school type assignments

>>62165011
This is why O-notation has the notions of "the worst case" and "amortized cost".

File: wows.jpg (9KB, 480x360px) Image search: [Google]

9KB, 480x360px

>>62164989
Dunno about Gatsby, but The Wolf of Wall Street is one of my favorite movies

>>62165063
Well fucking done, saying all that now, after I've done tests. It's incorrect to label the code as O(n^2) nor as O(n ln n).

>>62164989
It expects you to do the additional reading too.
A good second course after that is csApp, has lectures too https://www.cs.cmu.edu/~213/schedule.html but you need to book to go with it.

https://github.com/libmir/dcompute

Even if I'm not gonna use it, mir is showing a lot of promise

>>62162967

Never.
It's way too fragmented.

>>62165011
>>62165085
You realise that the big-O notation is about the worst case, right? Coefficients are ignored. If you have two algorithms that are O(n^2), it doesn't mean they will run at the same speed.

The only thing your measurements are showing is that you don't know how the notation is used, sorry.

>>62165163
Why? Only 2 versions.

Compared to something like Lua, which is currently three actively used separate languages, it's very unison.

>>62165085
>>62165203
I forgot to say: it's correct to label the code as O(n^2) because it's just that if you follow the rules of the notation.

Should I include header files in my makefile as dependencies?

>>62165226
Do you have code in your header files?

>>62165203
>You realise that the big-O notation is about the worst case, right?
Why wouldn't anyone include worst case and average case big-O ?

>>62165249
Yes.

>>62165226
You can, but you'll have to keep it in sync with your "#include"s manually, something you will most likely forget to do. I personally just do `make clean all` every time I change a header.

>>62165256
Then I would definitively consider adding them, yes.

How do I go from knowing the basics to making actual full fledged projects? I've read about data structures a bit but I can't seem to understand when and why to actually apply them. Any ideas for how to go from understanding syntax to actually creating something cool? So far the most complex thing i've worked on is probably the game of life

>>62163283
and there doesn't need to be any

>>62165271
>>62165279
What about listing a compiler as a dependency? Not a joke, since I have three generator scripts that are still work in progress.

>>62165255
Because `average` is too vague, no one knows that your average data is gonna look like, `amortized` is close, but not exactly.

>>62165284
Big project management is a skill that takes years of hard work to develop. Simply make bigger and bigger applications.

>>62165298
Makes sense if you need to rerun the scripts every time they change. Make itself has no notion of "compiler", it doesn't care what you're using your dependencies for.

>>62165284
Write features for open source projects, and then debug any problems that arise from said features. Repeat. It's the only way to learn.
Watch this: https://youtu.be/dvwkaHBrDyI

There's a few courses out there that specifically teach 'big project' understanding like these https://www.edx.org/micromasters/software-development (they are free, $$ is only for bs certs) or just dive into some open source project yourself.

Otherwise just make toy projects, like a web browser, Virtual Machine, ect. I started by making toy projects with limited functionality

Is there an elegant way to use std::visit?

>>62165363
std::come_in

>>62165373
std::hello

>>62165338

>web browser
>Virtual Machine
>toy projects


Hey Carmack, how are you doing?

>>62165363

CppWeekly - Ep 49 and Ep 50

>>62165385
std::fuck_you_john_you_fucked_my_wife

>>62165338

> Web Browser
> toy project

File: IMG_20170629_061812.jpg (2MB, 2448x3264px) Image search: [Google]

2MB, 2448x3264px

>>62165284
Also, if you want a general strategy for designing a largish program by yourself, go on libgen.io and get The Art of Computer Programming Vol1. Skip to page 191 (Or end of Chapter 1.4.1) where Knuth lays out step by step how to do this by breaking components into smaller ones, then starting with the smallest component and working your way up. Of course we have bullshit 'Agile' and other methods these days, but by yourself this works fine.

>>62165401
std::i_am_sorry_roger_it_just_kinda_happened

>>62165363
http://en.cppreference.com/w/cpp/utility/variant/visit

        // type-matching visitor: can also be a class with 4 overloaded operator()'s
        std::visit([](auto&& arg) {
            using T = std::decay_t<decltype(arg)>;
            if constexpr (std::is_same_v<T, int>)
                std::cout << "int with value " << arg << '\n';
            else if constexpr (std::is_same_v<T, long>)
                std::cout << "long with value " << arg << '\n';
            else if constexpr (std::is_same_v<T, double>)
                std::cout << "double with value " << arg << '\n';
            else if constexpr (std::is_same_v<T, std::string>)
                std::cout << "std::string with value " << std::quoted(arg) << '\n';
            else 
                static_assert(always_false<T>::value, "non-exhaustive visitor!");
        }, w);

>>62165415
std::you_have<1>

Explain this: http://austingroupbugs.net/view.php?id=523

Why can't I find .PHONY in the openstd POSIX 2016?

>>62165441
This is pretty clumsy-looking.

>>62165475
Of course it is, C++ has no pattern matching.

>>62165514
t. butthurt Rust shill

>>62165271
What if you listed all headers as a dependency of all? Then make would effectively do what you do manually automatically. You would only need to change the makefile when:

>deleting headers
>creating Headers
>renaming headers

but you already have to do this with source files.

>>62165516
It's not I who's butthurt tho.

>>62165514
I suppose you could also do something like this.

std::variant<int, double, char> a = 10, b = 10.4, c = 'x';
    struct 
    { 
        auto operator()(int i) { std::cout << "Called on int" << std::endl; };
        auto operator()(double d) { std::cout << "Called on double" << std::endl; };
        auto operator()(char c) { std::cout << "Called on char" << std::endl; };
    } visitor;
    std::visit(visitor, a);

That seems a little nicer.

>>62165516
Rust enums are lovely though. I'm mostly a sepplesfag.

>>62165553
Yeah, it is. You got rekt in the discussion yesterday and your ass must still be sore from all the pounding it received, since you're still all upset about it.

>>62165284

Look up "software engineering".

But for a single person, you only need very few tools. Agile and TDD is nice because you can adapt it for stand-alone programmers.

>>62165560
>rust enums
>rust traits
>rust results and options

>the least original language is now renaming everything
fucks sakes

>>62165580
>Rust is shit
>Rust has things I like
>Why are you talking about Rust?
???

>>62165514

> muh syntax sugar

kys

>>62165550
It should work, you can even put something like *.h in dependencies.
>>62165576
>there's only one Rust shill here
I have some bad news for you.

>>62165596
Are you replying to the wrong post?

File: 1417927240919.jpg (200KB, 600x443px) Image search: [Google]

200KB, 600x443px

>>62163877
Can you stip spamming this shitpost please? Noone cares about your shitty currency stoner

>>62165607
>pretending to be a different person when it's clear as day you're the same bloke as from yesterday
lol, absolutely pathetic

>>62165467
So is .PHONY standard or just a GNU make extension?

how does a lisp interpreter differ from a virtual environment like the jve?

>>62163147
that sounds like a game for little children

>>62165600
hll's are basically just syntactic sugar for assembly though

nice

> make 2>/dev/null
gcc -c -o ./main.o -std=c11 -Wall -Wextra -Wpedantic -O2 ./main.c
gcc -c -o ./g.o -std=c11 -Wall -Wextra -Wpedantic -O2 ./g.c
gcc -c -o ./tick.o -std=c11 -Wall -Wextra -Wpedantic -O2 ./tick.c
gcc -c -o ./draw.o -std=c11 -Wall -Wextra -Wpedantic -O2 ./draw.c
gcc -c -o ./dg.o -std=c11 -Wall -Wextra -Wpedantic -O2 ./dg.c
util/lvlgen.py util/rsc/0.cfg util/rsc/0.ppm rsc/0.lvl
util/flgen.py util/rsc/0.cfg util/rsc/0.ppm rsc/0.frames
util/sflgen.py util/rsc/0.cfg util/rsc/0.ppm rsc/0s.frames
gcc -o tabout -s ./main.o ./g.o ./tick.o ./draw.o ./dg.o -lm -lGLESv2 -lglfw
> ./tabout
Segmentation fault (core dumped)

>>62165982
That's progress.

File: Comfy_guy.jpg (53KB, 197x190px) Image search: [Google]

53KB, 197x190px

>CS
>first day C course
>done all assignments for two weeks ahead
>already working, alone, on group assignment that is supposed to be in in a month and will have it done tomorrow
ez life

>>62166055
>being *that* guy in class
I know that feel.

>>62166055
>Software Engineering
>C#
>Dropped out after first semester because was working on my game in C
>Got 1 year student holiday
>Resume on January
>Maybe I'll finish the fucking game before that

>>62166080
What game?
Genre?

>>62166080
btw university uses
>Microsoft Windows 10
>Microsoft Office 365
>Microsoft Outlook
>Microsoft Visual Studio 2011

>>62166098
Stupid 2D platformer. Also writing a game engine for it. I hope to extend it to 3D one day. Very bareback, links directly to GLFW (one step from direct X11 server access), libsoundio (one step from direct ALSA/PulseAudio/Jack access), GLESv2 and nuklear.

>>62166055
bet it must have felt good to implement that F to C converter, you've been training for this all your life

File: mylife.gif (152KB, 400x267px) Image search: [Google]

152KB, 400x267px

Why am I so bad a programming?

File: tmp_12407-1488961674122-1155081001.png (412KB, 1057x717px) Image search: [Google]

412KB, 1057x717px

So, is Go the server language?

>>62166132
>Microsoft Visual Studio
Thankfully my uni allowed me to install dev C++ on school computers. I can't stand MSVS

>>62166177
Dev C++ over MSVS
WTF!?

>>62166177
I begged for Code::Blocks and they said no.

>>62166177
How retarded are you?
t. used dev-cpp back in 2003

>>62166193
>>62166214
I just feel like it's much more cleaner and simpler if you're only working with C.

exceptions a shit

>>62166278
It's a complete piece of shit. If you want something clean and simple use gcc, gdb and some neat code editor or VS Code or Atom or even Geany anything goes as long as it's not dev-cpp :<

File: code::blocks.png (67KB, 440x190px) Image search: [Google]

67KB, 440x190px

>>62166344
Well, I'm going to shill my solution as well

>>62166405
btw compile from source latest svn checkout

>>62165408
Well suckless surf is something like 3k lines of C.
You are going to build it on top of webkit anyways so it's not that hard.

Why is POSIX make so ridiculously limited?

>>62166158
Yes.

>>62166158
No.

Do any of you guys find shell to be unlearnable? I learned C, Lua, Python and make quite easily, but never quite got around shell.

>>62166158
Maybe.

>>62166158
I don't know.

>>62166681
It's shit. Learn Perl.

>>62166746
What does Perl do that Python already doesn't?

>>62166746
hope you don't mean perl6. read a guy's follow up blog post to one where he said how excited he was to learn it and what i've gleaned is that it's a shitshow, and the only cool thing about it is that it's easy to extract command line options when writing a command line tool

>>62166761
the shit that you might want to write a shell script for

>>62166761
It is the other way round, y'know?

File: 1503659746434.jpg (44KB, 640x360px) Image search: [Google]

44KB, 640x360px

Can you repeat the question?

>>62166746
> echo the year is $((1995+22))
> wasting your time learning perl

>>62166772
perl a shit learn ruby

>>62166790
>>62166801
Isn't that what's the shutil module is for?

>>62166809
>learn perl-inspired language
I'm fine with it.

>>62166803
Why do you use malloc instead of calloc?

>>62166809
>>62166820
Are both an unreadable mess no one dares to touch?

>>62166831
callloc is for niggers

File: 1504047056680.png (311KB, 599x750px) Image search: [Google]

311KB, 599x750px

>>62166833
It's readable

It's 2017 and still
>building projects sucks
>distributing native libraries sucks
>cross compiling sucks
>android GUI applications have to be java programs instead letting you just create GUI with native executable
Who's fault is it that we still don't have anything nice?

>>62167087
Mine

>>62167087
You're autistic. 2017 is a great year for software development.

>>62167178
t. shitter who writes webapps

Posted from a web app.

>>62167087
᠋>distributing native libraries sucks
Why?

>>62167178
not even him but i'm autistic fuck off

>>62167244
You're a burden to society. Kill yourself.

>>62167239
>there are more package managers than atoms in the universe

>>62167250
if everyone who's a burden to society killed themselves there'd be no society left
>inb4 edgy bullshit like "that's what we should be aiming for"
objectively wrong, anything that doesn't pose a net benefit to humanity cannot under any circumstances be an object of "should"; there is no greater moral or utilitarian scope than the benefit of humanity

>>62162915
It would be better if the western society defended the rights of the actual disabled people than healthy gay faggots and pajeet shitskins.

>>62167087
>building
>distributing
>cross compiling
Rust solves all of those.

>>62167287
You shouldn't have the right to live if you're a burden to society. With burden I mean: you cost more than you bring in.

>>62167317
first of all: it would be better if the western society defended equally the rights of absolutely all parties whose rights are or have ever been under attack regardless of reason or disability status
secondly, >>>/pol/

>>62167332
rust is shit and not even a programming language.
Go jerk about your CoC at >>>/hackernews/ or >>>/github/

>>62167349
>With burden I mean: you cost more than you bring in.
oh
well then you're wrong about me
(i thought you meant any nonzero cost regardless of offset by deed)

>>62167371
If that's the case, I have no problem with that. I was talking about autistic NEETs living from autism bucks.

All my CS professors have horrible management skills. Half the time, they use blackboard, half the time they use their own personal homepages to assign homework. My theory of computation professor uploaded the assignment instructions to his homepage, and then uploaded the resources for that assignment (.tex files) to blackboard.

>>62162915
Lain opening is now in my head

>>62167477
self taught master race

>>62167477
dang, lol

>>62167522
>unemployed master race

>>62167522
I'm self taught in programming but skipped basically all of the theory stuff. Either way, I need the degree to get a job.

https://typelevel.org/cats/

>>62167477
you cant upload files to a blackboard, that's ridiculous

New thread
>>62167566

New thread:

>>62167573
>>62167573
>>62167573

Let's say I want to transpile an statically-linked x86 binaries to arm. Can I write the software to do that from scratch without going insane?
How much harder would doing the same thing for dynamically linked software be?

>>62167582
Nah bro. See here:
>>62167578

>>62167399
>I was talking about autistic NEETs living from autism bucks.
oh okay

that's not what autistic means though

autistic means possessing of autism
how profoundly it affects the displayed capacity of the afflicted to function in society is in no way suggested by the word

also although i go to school and intend to seek employment when i complete my education (i know my limits and i'm well aware i'm not quite high functioning enough to do both at once) it seems i have more sympathy than you do for those more limited than myself

i disagree with the premise that people who fundamentally lack the capacity--neurodevelopmental or otherwise--to reliably provide for their own lives in this society deserve to die

i think it's no one's fault if they do die for that reason

and i think no one could have been expected or obligated to take care of them

but i also think, even in the absence of any possible blame, their resulting death is still wrong in and of itself

wrong of whom you might ask

of no one

it's just wrong

it should not have happened

they should not have been born unable to care for themselves, nor made so by any crippling incident

it's wrong that that happened to them, it's wrong that it killed them, and they deserved better

and the fact that they deserved better is an immutable moral fact independent of the equally immutable moral fact that it was no one's job to help them and no one's fault that no one did

>>62167578
>discord
Never

cant we just have a rolling thread so this shit doesnt happen

>>62167619
What purpose does a cripple have? Besides being a burden to society, they're a burden to themselves. They know they cannot achieve all great things other can achieve. Living for the sake of being alive is stupid. You live to do great things that make you and the people around you proud.

>>62163147
>Even if it's not your personal cup of tea, you can't deny it's an excellent way of visually showing how programming work.

>Kid can't image what his line of code is doing where, or the overall project.

Good luck with your start up

File: ngpcnB0.jpg (391KB, 600x847px) Image search: [Google]

391KB, 600x847px

How do i get into systems programming on linux, the only skills i have are C and OpenGL

>>62167890
Learn how to compile linux. Learn how to make modules. Fix stuff in the kernel. Get insulted by Linus.

File: tumblr_oncis6mnZo1ts9rljo3_500.jpg (31KB, 480x480px) Image search: [Google]

31KB, 480x480px

>>62167916
Any good places to start?

>>62167961
Gentoo wiki for compiling. There used to be a thing called The Euduptyla challenge but I'm not sure it exists anymore.