/dpt/ - Daily Programming Thread

>>61750725
make a twitter bot that posts random comments from /dpt/ every few minutes

Second for C#

I'm currently reading this: http://yogthos.github.io/ClojureDistilled.html.
After that I'll be conjuring an RSS downloader to automate downloading torrents.

>>61750725
Ethereum miner in Rust

>>61750739
fuck

File: KinZiWJ.jpg (859KB, 2800x1925px) Image search: [Google]

859KB, 2800x1925px

I am getting into programming for the first time. I tried Python once around 6-7 years ago and found it to be easy but didn't continue because I got a job. Now I want to restart and I think I will start with Ruby because it seems to be even easier. Thoughts?

>>61750755
Java/C#?

>>61750755
>I think I will start with Ruby because it seems to be even easier
I don't think it's easier, in fact I think its weird syntax can be harder to learn
>>61750778
C# is great

>>61750755
you should start with java either way,

http://mooc.fi/english.html

do these and then you can move to ruby or whatever you want

>>61750755
>choosing a programming language based on perceived difficulty

if you want to learn something that is guaranteed to be useful for at least the next 10 years learn C++ or JavaScript

File: s.png (159KB, 1680x1050px) Image search: [Google]

159KB, 1680x1050px

>>61750708
>What are you working on, /g/?

Exploring the deepest, darkest corners of Racket.

>>61750786
>C# is great
It is, but I prefer Java because I prefer to stay away from the Microsoft world.

>>61750805
I wanted to say something nice to you, but then I saw your name.

>>61750755
I love ruby, I find it so much easier and fun than python. So for those higher languages, for me ruby is much better. I dislike Rails and am sad that ruby got so intimately attached to that junkpile, because I think as a language it has some things to offer.

No real difference though, so if you're already familiar with Python you could stick with it. Also, after you're exposed to a certain number of languages, a lot of them in certain "families" are basically the same except for some syntax.
I would suggest doing python or ruby, and also some other language, such as haskell or C. C will teach you about how low-level stuff works (and learning the actual parts that make up C is very quick) and haskell will teach you about functional programming, which is the present and future.

>>61750805
>#:key, #:when
explain this

How do I into embedded programming/dealing with embedded in general?

thanks anons

>>61750908
grab an embedded system and get cracking

>>61750805
I don't see the dark, deep corners you spoke of. This is some pretty standard (and fucking ugly on your part) sequence processing.
>>61750890
Some overengineered bullshit from racket's for/* macros. Not too unlike Common Lisp.

>>61750919
emacs in emacs

File: 123456yhjn.png (37KB, 509x659px) Image search: [Google]

37KB, 509x659px

>>61750919
>emacs

sublime text 3 master race

multithreaded emacs when?

>>61750708
Learning Rust. I need to think of a reasonably quick, possibly useful project to do with it when I have halfway-decent knowledge. I might make a clipboard editor, possibly with vim syntax.

>>61750954
but are there any good resources/good stuff to get/specific stuff to get started with

Full disclaimer: This is for the current sha2017 CTF.

Anyway. Consider the statement

eval(raw_input().title())

How do I e.g., make function calls when every word gets capitalized? (E.g., "asd" becomes "Asd". In addition, "ASD" also becomes "Asd".)

>>61750977
Does that lint-as-you-type, support refactorings, mark all syntax errors, etc?

Because if not, it's strictly inferior to Emacs.

>>61750919
gtk is better.

>tfw IDE-baby

intellij is so good tho, and the entire suite matches eachother (php storm, etc)

>>61750890
keyword arguments

What are some less known ways to make programming in C easier?
I know openMP but posix thread is more flexible.
I know compiler extensions but they are not really portable.
I could implement my own preprocessor but I don't really want to do that.

>>61751013
>lint-as-you-type
I actually find that annoying and prefer the occasional lint-when-you-save

>>61751060
don't worry, you can have that instead on emacs

>>61750708

Learning elisp.

>>61751089
I already have that on Vim though
fitemem8

>>61751012

capitalize = lambda s: s[0].upper()+s[1:].lower()
#testing...
print (capitalize("ASD")) # outputs Asd
print (capitalize("asd")) # outputs Asd

>>61750919

Why would you use emacs terminal? I understand vim > gvim, but why emacs?

>>61751121
reverse("emacs") ==> "scame"

>>61751100
how about a web browser in your editor?

>>61751147
why the fuck would i want a bloated editor?

why not make an electron app out of it?

idiot

>>61751147
http://www.vim.org/scripts/script.php?script_id=4315

>IDE build on top of web browser
>embedding buffer for external browser
>full circle

>>61751147
That sounds nice, I occasionally want to sendmail while watching Conway's game of life unfolding in my editor too.
How about no.

>>61751054
>tfw I choose my languages based on whether there is an editor/IDE existent that has auto-completion, go to definition, etc for it

REALLY stupid question but I'm not sure how to phrase it on google and it's faster to ask here: is it possible to initialise several variables at once? something like

int i,j,k = 42;
//now i == j == k == 42

>>61751175

int i, j, k;
i = j = k = 42;

This is the closest you can get

>>61751175

int i = 42, j = 42, k = 42;

>>61751116
I think you misunderstood me.

The point is, I input a string, which get "titled" and then evaluted. For example: If I input

print 123

The code that will be run is actually

Print 123

which will signal an error since

 Print 

isn't a function/statement in python.

I guess the TL;DR is: How do you circumvent the case sensitiveness of python?

>>61751184
Oh yeah nice, dunno why I didn't think of that

>>61751191
You're duplicating too much code

File: 1501111165081.png (14KB, 352x183px) Image search: [Google]

14KB, 352x183px

>>61751147
>yfw javascript interpreter included

>>61751147
>tfw vscode and atom
You are like little baby, my editor IS a web browser
That can't browse the web, but still

>>61751155
>why would I want a program that I spend a lot of time in do multiple things?
>>61751157
>only text, no images
>>61751169
you actually can send mail, and probably someone made Conway's game of life for emacs. You can make it anyway if no one did, that's the real beauty of emacs.

>>61751224
>you actually can send mail, and probably someone made Conway's game of life for emacs. You can make it anyway if no one did, that's the real beauty of emacs.
No, I said that because they were the classical hallmark of emacs' bloatedness.
So how about no.

emacs is contrary to the unix philosophy

>>61751216
>having an editor with google analytics built-in

Lol

>>61751248
>whole emacs is like 70mb
>meawhile any bigger IDE does not provide as much functionality but still takes way more space

>>61751268
Emacs was never rooted in the Unix culture though It's one of the leftover living fossil of the old Lisp culture.
That is to say, bloated.

>>61751224
>>61751278
I already have the shell to do all that for me. It even comes with its own language and command line!

Using my shell, I can run a text editor, a web browser, an email client, and Conway's game of life. Now, tell me, why would I want one of those to ALSO be able to run all the others from within itself? That's just bloat for no reason.

>why would I want a program that I spend a lot of time in do multiple things?
But I do, hence I like that I can do all those things from my shell. With tmux, I can even do many of those side-by-side.

>>61751298
Also, vim has access to the shell.

>>61751298
You can't have IDE features in your shell, or you can but it would be mighty awkward

>>61751298
but you can use only the one that runs all other other from within itself, just like you can only use the shell. It also provides greater functionality, as it has the ability to draw images and is easily extendable.

>>61751313
Checkout vim-slime my anon, I'm using it with tmux too.

>>61751318
>compiler
>syntax checker
>print the entire file
>everything

Yes, you can.

It's just:

:wq
*some shell task*
vim file.c

Is there any good guide to write an irc client, with all the things I need to implement, the protocol used, that doesn't require me to read an RFC?

>>61751336
>vim file.c
>invoking bloated shit just so you can pretend your shell is enough

>>61751336
>wq
Actually you can just :sh
Or

:!rm -rf /

>>61751268

It's just a lisp interpreter really.

>>61751298
Tmux and vim is a good combination. Used them for a long time before trying Emacs but now I'm hooked.

What sold me on emacs with Evil mode (vim hotkeys) is just how good the REPL integration is for python and Scheme. Also jumping around the shell with real vim hotkeys and not just the xterm vi mode is fantastic.

>>61751375
WHAT THE FUCK


WHERE ARE MY FILES

>>61751318
Why would I want an IDE in my shell? I have my text editor for that.

See why it makes no sense to run everything from your text editor? I don't need an IDE when reading emails or playing the game of life.

>>61751381
Fun fact: because accidental

rm -rf /

was an actual and somewhat common thing in the past, implementations of rm now all have failsafe check for this specific case.

Good projects for C beginners? I am trying to get my programming juices flowing. Thanks in advance.

>>61751390
>I don't need an IDE when reading emails or playing the game of life
usually, but if the situation occurs one day, why not?

>>61751401
write an OS

``Every program attempts to expand until it can read mail'' - jwz Zawinski

>>61751401
Write a particle system simulator implementing a 1/r^2 attraction rule and collisions. Use some package to visualize the simulation

>>61751424
On it. Sounds like something I could over my lunch break. Thanks!

>>61751175
In what language?
Closest thing you can get in Java is

int i = 1, a = 2,  b = 4;

But I'm sure that if you ask Hejlsberg, it'll be a feature in the next version of C#, where you can do int a,b,c = 4;

>>61751404
Why would the situation occur one day? That's inane.

If I read an email and want to have IDE features in its text, or if I want to write an email using IDE features, I'll use my text editor for that.

>>61751445
Dude, I did say I was a beginner, but goddamn, even a goddamn toddler could do this. Don't waste my time with shit that appears on the first page of introduction books.

>>61751204
Stictly speaking, you are duplicating too much code, since you're creating a bunch of vars that hold the same information.

>>61751445
This is impossible.

>>61751445
Different anon, that sounds fun! I might try that.

>>61751401
Tarjan's algorith
Read a tgf file and output either another tgf file of its condensation, or a graphviz file for easy visualisation

>>61751424
Joke aside can you even write a bootloader, lest an OS, in C alone these days, since the mainstream architectures now include all kind of bullshit that must be accessed by asm.

>>61751476
Set up a basic webpage using an ORM to pull some data from a database, and render it on a web page using a template.

>>61751489
Optional: use a priority queue to simulate it efficiently

>>61751445
>>61751476
Jokes aside, I would really like to see someone implement this, I am not good enough to do it myself. I will keep it on hold once I git guder.

studying for the GRE still. currently learning about 30-60-90 triangles. really didn't appreciate how applicable this is to CSCI until now

>>61751560
>GRE
?

>>61751534
I think you'll be surprised how simple it ends up being. The difficult part is really getting the math, right. Give it a try, honestly.

>>61751588
Thank you, anon. Programming has made math/physics more interesting for me.

>>61751588
how would you even start something like this?


t. noob

>>61751638
Just find the visualize library first and play around with it.
Then get to the math.
Then find a way to ram the math into the visualization.

>>61751560
Oh it's you from the other day. Yep learned that one in middle school too. Your high school diploma really should be revoked.

>>61751695
are you the same guy who posted the article about how highschool doesn't matter because highschool valedictorians do terrible in life? if so, that's ironic

>scalac takes 20+ minutes to compile a sub-1kloc project
wew

>>61751445
>1/r^2 attraction rule and collisions
>collisions
Come on anon at least use:
https://en.wikipedia.org/wiki/Lennard-Jones_potential

just realized i have no idea how regular expressions are implemented

would love to learn about this rather than doing GRE practice problems

File: 000069-newface.jpg (5000×5000).png (245KB, 1064x1069px) Image search: [Google]

245KB, 1064x1069px

>>61751795
hopefully there isnt too much math involved

File: fat.gif (2MB, 700x550px) Image search: [Google]

2MB, 700x550px

>>61747172
Does this count?

>>61751827
why would it involve math?

>>61750708
I don't know if this is the right place to ask, but I'm thinking of learning coding via a sort of "boot camp" program. I know a tiny bit of Java and Python but am interested in learning more. Is there much of a market for this sort of thing, and if anyone's been to one of these coding programs, was it any good?

>>61751638
F = ma
a_t2 = (v_t2-v_t1)/(t2-t1), for example.
Solve for v_t2 at each time step. For each particle. Resolve any collisions by setting v = -v for each particle in the collision. Contain the system in a box, either by letting particles teleport from one side of the box to the other, or by having particles collide with the sides.

Now, code!

>>61751845
Regular expression is something comes straight outa computation theory textbook.
May you even finally understand why using regex to ``parse'' html or xml is a bad idea.

>>61751795
finite state machines or automata

File: Screen Shot 2017-08-05 at 4.51.18 PM.png (21KB, 700x550px) Image search: [Google]

21KB, 700x550px

>>61751830

>>61751902
what are some language theory textbooks that i can read to learn about regular context free and all those other grammars + regex on the side.

Working on making a Linux based OS right now

>>61751830
this is quite pleasant

>>61751887
Coding boot camps are shit and rightfully dying out. They're essentially outsourced job training. Most people who go to them already know how to program and were sent there by the company they just got a job at.

>>61751932
Because when water falls, it's not immediately sucked in by any nearby hole.

>>61751932
Is that an animated png? The thumbnail color looks right but I swear the full size image has purple water.

>>61751932
From what I can gather on this is image is that the water stops "flowing" on the area once it finds an exit. It happens again on the second to last area. Not the guy who made this. It seems to be a simple over sight.

>>61751902
computation theory isn't math firstly, and secondly i'm talking about how they're implemented in perl

>>61751989
I see. So it probably wouldn't be ideal for someone who's seeking a job?

>>61751993
None of the other containers exhibit that behavior, as you can see

>>61751958
The current standard textbook is Sipser's if I'm not wrong. The classic is Alfred V.Aho's though.
>>61752022
>computation theory isn't math firstly
Except it is
>b-but it's not like my highschool algebra and precalc

>>61752026
jesus christ no. Stay the hell away. Your money would be better spent hiring a private tutor to supplement your self-study.
You need a portfolio for job-seeking, and to actually know your shit. You can't just drop a ton of knowledge on someone in a month and consider them to understand it. Bootcamps are a huge scam.

>>61752040
Come on anon didn't you know that that proofs aren't math? Mrs P. didn't teach me any proofs in arithmetic class, so words aren't math. Only numbers are math.

Also thanks for book recommendations.

>>61752022
You should learn how the vanilla regex works first anyway. Perl's implementation extends regex power quite a lot, but it's not something entirely different.

>>61752035
Of course none of the other have that behavior because they don't meet the same conditions.

The protrusion is at the same height as a hole that creates a flow in the opposite direction.

>>61752040
dude, i've taken the automata class. it's not the same thing as doing a bunch of algebra and geometry for the GRE

fucking math majors

>>61752069
Well, that's a shame. Thank you for the warning, though. I'll keep looking around.

>>61750737
twitter.com/dpttxt is already doing that task for a pool of dpt shitposts

hate math majors on /dpt/. they're insufferable, constantly playing dumb. the exact same type of person to love stefan molyneux and listen to game of throne ebooks

>>61752076
Well actually I think if you're on a roll, get some compiler textbook while you're at it too. They cover finite and push down automata and have actual application for them too, because you know, lexer and parser, and who in their right mind doesn't want to write a toy compiler at least once in their life?

>math major

jokes on you, i dropped out of mechanical engineering in final year so i don't actually have a degree

File: iu[1].jpg (22KB, 260x335px) Image search: [Google]

22KB, 260x335px

>>61752150
>>61752076
>>61752040
>>61751958
actually now that i think about it, this is a pretty good resource for computer science students too. you really shouldn't learn about regex until you've reviewed everything you've learned in highschool

>>61750805
>>61750959

Why the fuck are macros still a thing? FEXPs are the real deal, just a function where you don't evaluate arguments (or the result, as a macro would. FEXPs are functionally equivalent to adding custom clauses to eval itself.

>>61752144
Honestly though how did you get through school without learning basic geometry? Seriously though anon. I have trouble believing that you weren't exposed to triangles in school.

>>61752207
i don't even post here mate, dunno what's going on haha

here's the context of why I'm here >>>/int/77992679

>>61752231
yesterday i posted that i wanted to keep learning assembly but i can't because i have the GRE upcoming. he's a twat who's a dead center on the IQ spectrum who thinks high GPA's are bad because he graduated the university of life. he's this guy >>61752169

>>61752199
>FEXPs are the real deal, just a function where you don't evaluate arguments
Aren't they just simply non strict functions then? The point with macros is to have a way to manipulate the AST during compile/eval time.

Can you use C++ with OpenGL? Or is it only/mainly focused on C?

>>61752248
*pushes you over and kicks your bent little head in*

haha nothing personnel you gaylord haha

File: apostol.jpg (14KB, 332x475px) Image search: [Google]

14KB, 332x475px

>>61752248
Lemme give you a recommendation to learn math.

>>61752270
If you can use it with C, you can use it with C++.

>>61752294
I know C++ decently well but not really C much.

>>61752292
my test's in 10 days

>>61752300
You should fix that.

>>61752300
i think either you don't know C++ as well as you think, or you know more C than you think without realizing it

>>61752255
You're given arguments as unevaluated lists. You can just change those and then evaluate them, or return them, or whatever.

Special function types like FEXPs are implemented in eval. There's no reason to think they wouldn't be available everywhere eval is available. While a regular function call is implemented as (apply f (eval args), FEXPs is just (apply f args).

>>61752332
wrong

>>61752270
Yeah you can but if you're deep in C++, you're going to not like it because OpenGL uses C idioms and doesn't have classes. I'm not aware of a C++ wrapper that makes OpenGL class based (although it wouldn't be hard to make one). I think lwjgl does that but it's Java.

Shit like this:

glGenQueries(2, time_queries);
glGenBuffers(1, &menu_vbo);
glBindBuffer(GL_ARRAY_BUFFER, menu_vbo);

>>61752199
macro are for intelligent code generation and are evaluated in the static environment.

>>61752332
So it's just a lazy lambda basically?

I guess it helps in some ways but macros contain syntax that's illegal for functions. Also most non-trival macros aren't simple functions.

Let

(let ((a 1)) a)

(()) ain't valid scheme.

Cond:

(cond (((member lst) => (lambda (x) (print (car x))))))

(()) again, and this time =>, which aint a variable.

True things like (if) could be implemented as an f expression no problem but most macros can't.

>>61752332
But are FEXPs available at compile time, because macros expansion is essentially compile time computation.
Otherwise your FEXPs are just simply non strict function. Haskell functions are non strict, and do not evaluate their arguments unless specifically need to

if' test ex1 ex2 = if test then ex1 else ex2

Yet it still needs Template Haskell, the equivalent to Lisp macros, to perform compile time AST manipulation.

Testing code that primarily runs at the type level means that the tests must also primarily run at the type level.

This means that the tests will probably run very quickly, but compile very slowly.

>>61752522
>type level computation == compile slowly
I-is this the "let's laugh at the scalafag" thread?

>>61752357
why

>>61752372
So what? If they're evaluated at compile time, that means eval is available during compilation, therefore the whole language must be available.

>>61752450
>(()) ain't valid scheme.

It's invalid if you evaluate it. ((a 1)) is just a list with another list (a 1) as it's car and nil as its cdr. Obviously it will fail if you try to evaluate it without compiling it into something else first. You can't apply a list.

>>61752474
>Template

Wot

That's the problem with you people, you fell into this "pattern-based templates" meme. It's just stupid in the most needlessly complicated manner possible. Scheme's syntax-case is a specially sad little construct.

You get a list of things instead of some value. You can change the list. You can evaluate the list. That's simple.

Daily reminder that every propositional formula can be converted into an equivalent formula in conjunctive normal form

>>61752450
>(()) ain't valid scheme.

> '((a 1))
=> ((a 1))

wot

File: 1463364999713.jpg (59KB, 268x237px) Image search: [Google]

59KB, 268x237px

Employed Haskell programmer here

>>61752655
how many times a day do you feel bad about the terrible type system you're subjected to?

>>61752595
I just can't believe how badly the compiler performs. Maybe if they took removed all the OO shit from the language then it would have a chance of running at a reasonable speed

>>61752621
Who even said a thing about the specific macros system of Scheme? I only brought up Template Haskell because it provides the same functionality as Lisp macros.
And it seems you still haven't understood that your Fexprs are just non strict. Lazy. Functions. They are a different beast and not a substitution for macros. See the if' implementation in Haskell again. Protip: it won't fucking diverge like in the SICP's my-if exercise.

File: tumblr_inline_no001bPsm71rq29x2_540.jpg (44KB, 450x618px) Image search: [Google]

44KB, 450x618px

Employed MUMPS programmer here

>>61752699
lol
http://thedailywtf.com/articles/A_Case_of_the_MUMPS

>>61752682
>Maybe if they took removed all the OO shit from the language then it would have a chance of running at a reasonable speed
Exactly. You know why? Because subtype relation are a fucking ordered relation instead of equality like other type relation. That's why other functional languages with expressive type system compile in a breeze but not fucking scala: the type checker/inferencer just has to do some unification and that's it.
With ordered type relation, the search space just explodes. EXPONENTIALLY.

>>61752621
Dude all scheme is lists

(+ 1 0)

is a list
I mean sure you could pollute your functions with quoted data but that's just nasty and there's something deeply disgusting about having to use eval to implement a simple let.

(define-syntax (syntax-rules () ((_ ((arg val) ...) body ...) ((lambda (arg ...) body ...) val ...))))
(let ((a (square 1))) a)

vs

(define-lazy (let args . body) (eval (list (list 'lambda (map car args) body) (map cdr args))))
(let `(a ,(square 1)) 'a)

Quaziquoting already feels dirty don't make me quaziquote everywhere due to syntactic closure.

The lazy let just feels fucking wrong. Macros allow you to write new languages based on scheme, but the fact that I have to quaziquote everything shatters the illusion.

>>61752743
s/are/is/
In other words, Scala just fucked itself in the get go.

>>61752743
I hate the OO and subtyping meme, it isn't funny any more

File: unnamed.jpg (36KB, 1256x534px) Image search: [Google]

36KB, 1256x534px

>>61752710
>no operator precedence
>one universal datatype
>Multiple statements per line are idiomatic. Scope of IF and FOR is "remainder of current line."

File: 1490498267810.jpg (248KB, 2048x1536px) Image search: [Google]

248KB, 2048x1536px

Employed neet reporting in

>>61752621
>So what? If they're evaluated at compile time, that means eval is available during compilation, therefore the whole language must be available.

A macro can't access the dynamic environment and is working on _syntax_ only. For a macro, 2 is not a number, it's only the character "2". A macro does not eval, it does expand. Can you do something like this with fexp?

(clang ;; expands C syntax into lisp syntax
  void foo(int x)
  {
    return x + 1;
  }
)

(foo 5) ;; -> 6

File: Selection_20170806_00:48:34.png (19KB, 613x108px) Image search: [Google]

19KB, 613x108px

>>61752773
MUMPS is widely used in hospitals to this very day

>>61750708
Does somebody have any book recommendations on test management subject?

>>61747172
G

>>61752800
>make a tool that converts MUMPS to a modern techstack
>license the technology to hospitals with the condition that I get to personally explain to MUMPS devs how their skillset is now completely worthless
I have a mission

Scheme or Clojure? Pros and cons of one over the other? I'm currently migrating from Haskell because I feel like it's not incredibly practical, but I do have a little experience with Scheme.

>>61752893
t. couldn't cope with haskell

>>61752621
Implement let* with your replacement for macros.
http://api.call-cc.org/doc/scheme/let%2A

>>61752907
Basically.

>>61752917
haskell's the standard for functional languages. if you don't know it and are sick of it, don't a lisp, learn php or javascript or something

>>61752754
>there's something deeply disgusting about having to use eval to implement a simple let.

As if macro is any better. A macro is just an FEXP that evaluates its own result. It's not "ugly" to you because you don't see it.

>Quaziquoting already feels dirty don't make me quaziquote everywhere due to syntactic closure.

FEXPs don't evaluate their arguments. Therefore they don't have to be passed in quoted. Have you even seen an FEXP

>>61752790
>A macro can't access the dynamic environment

That's a stupid limitation born by stupid things like dynamic scoping, which got fixed in Scheme. In fact that was the big argument against FEXPs back in the day. FEXPs were "too powerful" they said, "you can shoot yourself in the foot by clobbering identifiers".

>is working on _syntax_ only

That's another stupid limitation.

>For a macro, 2 is not a number, it's only the character "2".

That's not really working on Lisp syntax, that's working on lists of characters.

>>61752943
Yeah no

imagine using a language without support for parallelism using threads

>>61752913
It'd be similar to how >>61752754 did it, but you wouldn't have to quote the arguments as he implied.

Here's if implemented in newLISP (define-macro actually defines an fexpr):

(define-macro (my-if test true-action false-action)
   (if (eval test) (eval true-action) (eval false-action)))

(my-if (> 3 2) (println "yes it is" ) (exit))

Extremely simple

>>61752743
OO confirmed for inherently shit performance

>>61753002
No, implement let*.

Implement it in newlisp.

what's the best type system
what's the language with the best type system

File: 1498406374792.png (208KB, 499x499px) Image search: [Google]

208KB, 499x499px

>>61752871
Create a 2nd tool that transpiles a modern language to MUMPS
Get a MUMPS job
Be 10x as effective but tell no one
Browse 4chan and post frogs all day while getting paid $110k

File: 13775726487.png (557KB, 650x800px) Image search: [Google]

557KB, 650x800px

>>61752893
clojure has more tools and library and is much more interesting for concurrency. clojure is also more versed into functional programming. the list abstract type is much less important in clojure; cons in clojure does not return a pair (eg: (cons 1 2) is illegal)

https://clojure.org/reference/lisps

if you want to try clojure. get this https://sekao.net/nightcode/

How are Scala compile times nowadays?

>>61752943
>haskell's the standard for functional languages
The standard lazy functional language, sure. The standard for functional language should be something like SML and Scheme, because you know, only those two actually have formally defined semantics.

>>61753070
Thank you!

>>61753078
i'm not saying what should be, i'm saying what is. ask someone to name a functional language in the real world or at your work

>>61753078
I'm glad you said SML and not OCaml

>>61753049
Idris

>>61753049
F*

>>61753076
If you're willing to wait around I might be able to tell you later in the thread

>>61752199
kent pitman debunked fexp 40 years ago

>>61753070
>cons in clojure does not return a pair (eg: (cons 1 2) is illegal)
>https://clojure.org/reference/lisps
Oh my fucking god why would anyone use that piece of shit.

>>61752943
>haskell's the standard for functional languages.
no

how do i link libraries i got from the internet while compiling with gcc? most libraries have only have makefiles and there is no instruction on what i should add on terminal commands and everything. everything on gcc's website seems too vague and all these libraries they all install themselves on different directories after getting make'd.

ps: newfag to c.

>>61753126
Good

>>61753156
Face it, OCaml a shit, Idris not yet ready, Lisp basically imperative

>>61753137
>muh dynamic scoping

Just because it didn't play well with some 80s Lisps doesn't mean it's not good.

(define (a) 1)
(define (alpha) (a))
(alpha)             ;  1
(define (a) -1)
(alpha)             ; -1

(define-syntax b (syntax-rules () ((_) 2)))
(define (beta) (b))
(beta)              ;  2
(define-syntax b (syntax-rules () ((_) -2)))
(beta)              ;  2 still!

>>61753048
I don't really know hot to implement let nor do I care to look it up. There's no reason it shouldn't be possible.

>>61753196
>There's no reason it shouldn't be possible.
Bet you think there's no reason it shouldn't be possible to solve the halting problem either

>>61753022
It's mosty a Scala thing though. The type system just mixed different ideas the wrong way. OCaml is both OO and FP and gets by just fine. Heck, even Java and C# mix parametric polymorphism (not higher kinded) with subtyping just fine, too. Maybe that's why Scala implementors are proposing to drop higher kinded polymorphism from the language. I'd rather drop subtyping tbqdesu.

>>61753209
If you can implement let with a macro then you can do it with a fexp.

>>61753228
>I'd rather drop subtyping tbqdesu.
Same.

We don't need another Java/C#/F#/OCaml

>>61753196
I'm talking about let* not let.

I chose let* because I'm 99% sure it's impossible to implement without giving up lexical scoping. And giving up lexical scoping is retarded.

I see nothing wrong with your code by the way.

>>61753241
It's not let, it's let*.

>>61753181
clojure, f#, erlang, scala, java8, elexir, kotlin, ... those are more relevant than haskell.

>>61753241
>If you can implement x at compile time by macro expansion then you can implement x at runtime by lazy function
The Sussman is disappoint

>>61753268
>let*

If you can implement let with a regular Scheme function, then you can do it with a FEXP, which is just a regular function that doesn't eval arguments.

No idea how it's done since I've never done it.

>I see nothing wrong with your code by the way.

The macro doesn't change when the things it depends on change. It's static, not dynamic.

>>61753309
Nothing stops you from expanding some "macro" at runtime

File: smug.jpg (11KB, 205x246px) Image search: [Google]

11KB, 205x246px

>>61753308
>kotlin
>functional

File: 1492805709960.png (86KB, 273x353px) Image search: [Google]

86KB, 273x353px

>>61753343
>Nothing stops you from expanding some "macro" at runtime
But everything stops Fexp from "functioning" at compile time. Unless, of course, if they were used inside of a macro.

>>61753308
We're talking about which language is standard, not most beloved by poor programmers.

>>61753392
>everything stops Fexp from "functioning" at compile time

wot

FEXPs are really just a special case in eval.

Can you eval at compile time? If yes, then you can use FEXP, if not, then your lisp is crap

Common Lisp doesn't even require macros to be compile time constructs

>>61753343
Run this in a scheme repl:

(define (foo a)
  (eval a))
(let ((square (lambda (x) (* x x)))) (foo '(square 2)))

See how it doesn't work? It doesn't work for good reason. There's a reason why the "let" I posted needed quaziquoting. And that same reason makes let* impossible in your eval based quoted data "f-expressions"

Please learn how lisps work before shitposting about them.

>>61753455
>quotes square
>expects another environment to be able to resolve it

??

>>61753434
>Can you eval at compile time?
Calling eval by itself already makes the current level of computation the "compile time" for the eval'ed sexpr. So no, your Fexpr is already parsed into an AST by eval by the time any further post-compilation computation occurred.
Seems like it's impossible for you to understand why non strict functions are not the same as macros, so I guess I'm just done. Have fun implementing the other dude's let*.

>>61753509
That code would run in a lisp with dynamic scoping. But dynamic scoping is retarded.

>>61753002
>newlisp
newlisp is crap, it doens't have pair (cons always returns a list) and had to implement macros because of how crap fexpr are.

>>61753524
>Calling eval by itself already makes the current level of computation the "compile time" for the eval'ed sexpr.

>calling eval
>compile time
kek

redpill me on cubical type theory

>>61753584
ut's one of the better productive lisps. the pair thing isnt a problem and macros can be implemented on top of fexpr anyway so why not?

>>61753626
>calling eval
>compile time
Yeah? Your lisp doesn't let you run arbitrary code during the macro expansion phase or the reader phase?

Wow it's shit sempai.

Patterns and templates are pretty great IMO, lets you implement a lot of looping and mapping very concisely. syntax-rules gets a bad rep because you can't call arbitrary Scheme functions. However Racket's syntax-parse is extremely powerful, lets you do shit that would be horrible with just c[a/d]r, map and quasiquote.

Right now I'm doing something like this:

enum RETURN_VALUES {
    SUCCESS = 0;
    ERROR_A = -1;
    ERROR_B = -2;
    ERROR_C = -3;
};

int function1 (T *var){
    if(condition_A)
        return ERROR_A;
    if(condition_B)
        return ERROR_B;
    //do stuff
    return SUCCESS;
}

int function2 (T *var){
    if(condition_A)
        return ERROR_A;
    if(condition_C)
        return ERROR_C;
    //do stuff
    return SUCCESS;
}

int function3 (T *var){
    if(condition_B)
        return ERROR_B;
    if(condition_C)
        return ERROR_C;
    //do stuff
    return SUCCESS;
}

int function4 (T *var){
    if(condition_A)
        return ERROR_A;
    if(condition_B)
        return ERROR_B;
    if(condition_C)
        return ERROR_C;
    //do stuff
    return SUCCESS;
}

Is there a smarter way of doing the error checking? Each function checks for different errors, but writing the error conditions in every function feels a bit dumb.

>>61753700
If the conditions are the same for both functions then just make a seperate function that checks the conditions

>>61753700
use asserts my nigga

#include <assert.h>

#define SUCCESS 0

int function1 (T *var)
{
    assert(condition_A);
    assert(condition_B);
    //do stuff
    return SUCCESS;
}

int function2 (T *var)
{
    assert(condition_A);
    assert(condition_C);
    //do stuff
    return SUCCESS;
}

int function3 (T *var){
    assert(condition_B);
    assert(condition_C);
    //do stuff
    return SUCCESS;
}

int function4 (T *var)
{
    assert(condition_A);
    assert(condition_B);
    assert(condition_C);
    //do stuff
    return SUCCESS;
}

in java whats the proper way to save/recall the original constructor variable values? for like a reinitialize method that restores the object back to those values.

>>61753700
Im not sure what language that is, but could you just not throw if you have some error?

>>61753732
Just save them as private variables. Enterprise solution: use the Prototype Pattern (tm)

If you had to write a sieve right now from memory, could you do it? Keep in mind the Sieve of Eratosthenes is a trivial algorithms (i.e. mark off multiples of primes), and as such if fizzbuzz-tier.

>>61753750
not really.

i mean, what's the point?

and what sieve?

>>61753729
lol

>>61753772
?

>>61753763
>i mean, what's the point?
Proving, at least to yourself, that you're not incompetent.

where's the /dpt/ github

>>61753763
get back to studying for the gre

>>61753786
Yeah because competence comes from memorizing arbitrary algorithms.

>>61753729
I've heard some people saying that using assert isn't quite good because usually it is not present in the release build.

>>61753739
It is C. I don't know if I can throw in C but I like returning things

>Rc<(RefCell<Bpasi>, RefCell<Vec<Vec<u32>>>)>;
So this is the power of Rust...

>>61753821
Rust is fine until you need refcounting and then the types become absurdly verbose

>>61753813
You can raise signals.

>>61753807
>memorizing
You should be able to come up with it yourself, most trivial algorithms you can. If you can't work your way to a working algorithm from "mark off multiples of primes", you're not a real programmer.

>>61753862
Don't. Returning integers is the correct way to do things in C

>>61753876
They have different uses baka.

>>61753700
That's exactly what robust error handling looks like in C

Enumerate every error condition and return status codes from every function that can fail

>>61753862
Please don't EVER do this
Signals are shit enough with just the OS throwing them at you

>>61753874
>you're not a real programmer.
i don't even program

>>61753876
>is the correct way to do things in C
>muh Cthonic
Stick to python faggot

>>61753626
>kek
Never having read SICP confirmed.

>>61753900
Stick to HTML bitch

>>61750708

(define (tree-map func tree)
  (cond ((null? tree) tree)
    ((not (pair? tree)) (func tree))
    (else (cons (tree-map func (car tree))
            (tree-map func (car tree))))))

(define tree-a (list 1 (list 2 (list 3 4) 5) (list 6 7)))

Why is my code returning (1 1)? I can't work it out.

>>61753918
>Never having read SICP confirmed.
what for? mental masturbation?

meanwhile there are people out there earning more thatn $100K a year and don't even know what SICP is

>>61753855
>x is fine until it is not

>>61753918
MIT doesn't even use that piece of shit anymore. Drop it already

>>61753932

(tree-map (lambda (x) (* x x))
      (list 1 (list 2 (list 3 4) 5) (list 6 7)))

missed off this part

>>61753900
"Who cares about uniformity and standards? Fucking python babbies"

Ironically, one of the reasons I dislike Python is because its users claim righteousness about "pythonic" and that stupid manifesto, but in reality the modules follow little convention and are horribly inconsistent.

>>61753937
>mfw my first job out of uni was 88k/year and all i did was maintain js and php sites

this was in florida so your mileage will differ

>>61753941
Yes, some things are not black and white, and some things have both strengths and weaknesses.

>>61753963

The Zen of Python, by Tim Peters

Beautiful is better than ugly.
Explicit is better than implicit.
Simple is better than complex.
Complex is better than complicated.
Flat is better than nested.
Sparse is better than dense.
Readability counts.
Special cases aren't special enough to break the rules.
Although practicality beats purity.
Errors should never pass silently.
Unless explicitly silenced.
In the face of ambiguity, refuse the temptation to guess.
There should be one-- and preferably only one --obvious way to do it.
Although that way may not be obvious at first unless you're Dutch.
Now is better than never.
Although never is often better than *right* now.
If the implementation is hard to explain, it's a bad idea.
If the implementation is easy to explain, it may be a good idea.
Namespaces are one honking great idea -- let's do more of those!

>>61753963

I remember reading one of the most disgusting functions I've ever seen in a project because the dev autistic self-bandaged himself with a 79 char line-width.

>>61753963
Signals aren't equivalent to returning ints anon.

You can shoot yourself in the foot with them but you should learn use them because otherwise a user pressing control-c is going corrupt your program's data.

How do you go from a shitty/mediocre programmer, to a decent/god-tier programmer?
Also, any tips on discipline? Ways to code on days that you don't feel motivated? How do you guys deal with that?

>>61754008
You can't. Accept that you're forever going to stay medium talent and find a job that suits it

>>61753992
Sure but you shouldn't be triggering signals to do error handling, like you suggested. This is just begging for memory leaks when you can't clean up from an error. This is why people use integer return values

>>61754008
Read code written by people better at it than you.

>>61754066
Are you better than me?

>>61754076
Yes.

>>61753169
bump for this

>>61754066
So, that works on programming? So, it is like learning songs off your guitar heroes to understand their "reasoning" behind every note/chord. That's a pretty good advice, mate. Thanks.

>>61754086
Damn.

Where can i see your code (github, bitbucket,...)?

>>61754092
I'm not sure if anybody knows how desu. I'm not sure if the linker authors know how to.

I end up just googling around like crazy to find the magic invocation I have to pass to the linker.

>>61753169
sudo make install
It should put files into /usr/share/include, which you can include with #include <HEADERNAME>. Then you can link by passing -lLIBNAME to the compiler, which will search for a file called libLIBNAME.so in the path /usr/local/lib. I had to do some configuration for this on arch, since apparently arch doesnt automatically look for this path by default. Let me know if you have questions about that

File: Screenshot_2017-08-05_17-30-30.png (19KB, 449x144px) Image search: [Google]

19KB, 449x144px

>The type above is a hint for those that don't convert the tree into a list before applying the folding function.
Is it even possible to write a fold for a bintree without flattening it with a monoid?
The only idea I have right now is replacing each leaf with the initial value, but that's semantically very different.
...Then again, it just says "write a catamorphism", which it technically is, and they say any traversal order is fine, so would that still be a correct answer?

this code won't work on my pc || gcc version 6.3.0 (MinGW.org GCC-6.3.0-1)

#include <vector>
#include <tuple>
#include <iostream>

int main()
{
    int numbers_a[] = {0, 2, 4, 5, 6, 8, 9};
    int numbers_b[] = {1, 3, 5, 7, 8};

    std::vector<std::tuple<int, int>> pairs;
    for (auto &a : numbers_a) {
        for (auto &b : numbers_b) {
            if (a < b) {
                pairs.push_back(std::make_tuple(a, b));
            }
        }
    }

    for (auto &[a, b] : pairs) {
        std::cout << a << " is less than " << b << std::endl;
    }

    return 0;
}

only if i change the last loop to

for (int i = 0; i != pairs.size(); i++) {
    std::cout << i << std::endl;
}

>>61754150
Thats overkill, just traverse all the leaves using recursion desu

>>61754161
Is that C++17 syntax? What version are you on?

>>61750708
Procedurally generated robot mmo that will never be finished.

I'm currently refactoring some shit I wrote a year ago, and it's bad enough I'm losing inspiration every time I look at it

>>61754136
Most libraries for me (debian) are in /usr/lib. The only things in /usr/local/lib are folders for python.

But even then /usr/lib is missing some of my libraries. So no idea where those got installed to.

>>61754192
g++ -v gives

Using built-in specs.
COLLECT_GCC=g++
COLLECT_LTO_WRAPPER=c:/mingw/bin/../libexec/gcc/mingw32/6.3.0/lto-wrapper.exe
Target: mingw32
Thread model: win32
gcc version 6.3.0 (MinGW.org GCC-6.3.0-1)

>>61754203
You need to just accept imperfection and move on anon. Part of software development involves battling your inner demons and accepting that a solution is workable, without going too far and letting the kludges accumulate.

Nothing will ever be perfect.

>>61754206
Libraries installed by your package managers go into /usr/lib, but when you build from source (make install) they should go into /usr/local/lib. Usually you can tell the makefile which folder to choose as well

>>61754163
I mean that's already kind of what I have in mind.

data Tree a =
    Leaf
  | Node (Tree a) a (Tree a)

foldt :: (a -> b -> b) -> b -> Tree a -> b
foldt f z (Node left a right) =
  let z' = a `f` foldt f z left
  in  foldt f z' right

But that leaves the question of the base case, and

foldt f z Leaf = z

is the only thing that makes any sense algorithmically.

>>61754208
I think the [a, b] syntax is called structured bindings, in which case you need gcc 7 according to this: http://en.cppreference.com/w/cpp/compiler_support

>>61754248
Wait, I have an idea using partial application for child nodes.

>>61754248
I don't see anything wrong with that

>>61754235
Yeah I found that the libraries exist in /usr/lib with "ls -R | grep lib{name}.so" but I have no idea what folder they're in desu. "ls -R /usr/lib" for me has got to be dozens and dozens of screens.

How am I supposed to know that if I #include "foo.h" I need to pass -lfubar? There's gotta be a better way than going through dozens of folders in /usr/lib

>>61754273
so how do i print the contents of pairs?

I'm working in c++ and I have a bit of an issue with #define #ifdef etc.
I'm defining something in my main.cpp file that contains my main() function and I want to check whether this is defined in one of my .cpp files however this cpp file seems to be compiled before my main.cpp file so my definition is never defined there

main.cpp

#pragma once
#define my_def
#include "myClass.h"
//other stuff

myClass.h

#pragma once
//my_def is defined here

class myClass{
   void foo();
};

myClass.cpp

#include "myClass.h"

void myClass::foo() {
#ifdef mydef
//A
#else
//B <- It will always go here since my_def is never defined here
#endif
}

>>61754302
including and linking are very different things. Usually when you want to use a library you should know which headers it uses AND which libraries you're supposed to link with. E.g. <chibi/eval.h> plus libchibi-scheme.so

>>61751056
get into code generators and domain specific languages (dsl). to get a feelin for it google 'xtext' or 'xpand'. you could also have glance at templating engines and build systems. it depends on your use case. do you want to implement network protocols on top of the tcp/ip stack or write multi-platform code? model-driven 'software engineering' could also be interesting for you. as i said, it depends on your use case and requirements.
i once had a job where i had to write code for tizen os (fuck that shit). i used a few python scripts i wrote by myself to generate the user interfaces (ETL fuck that too) from a statechart given in scxml and a description of the ui (as containers). worked for me because my boss was like 'we prototype and if it feels good we'll use it'. we stopped the project because tizen was to unstable back then (and everyone in the western world uses android or ios, only asians use tizen but samsung keeps pushing it for their shitty smart-whatever platforms).

>>61754311
You're using <tuple>, not <pair>.

for (auto &p : pairs) {
  std::cout << std::get<0>(p) << " is less than " << std::get<1>(p) << std::endl; 
}

So reading this:
http://highscalability.com/blog/2013/5/13/the-secret-to-10-million-concurrent-connections-the-kernel-i.html

Is there any open drivers I could find and use that actually lets my servers go _FAST_?

>>61754323
>Usually
Thats' the magic word though.

Sometimes it's really not obvious. Is the library libfubar.so or is it libFUBAR.so or is it libfubar8.so or what? I have yet to find documentation for any library that I've used which says "the name of the library is libFUbar.so"

Sometimes you can guess but when it's something insane it's fucking horrible.

>>61754343
*EFL enlightment foundation library

>>61754359
god bless anon

>>61754366
Huh. In my experience it has been pretty straightforward, but I guess badly documented libraries are just that; badly documented.

File: 1500764358365.jpg (159KB, 750x864px) Image search: [Google]

159KB, 750x864px

I've been learning C++ for a while now and was wondering what kind of jobs I can apply for that aren't really Wall Street/Financial sector or gaming related? I have zero experience and find the number of years of experience all of the jobs require daunting.

Any help at all would be appreciated as I feel really lost.

>>61754410
People posting job listings with high requirements are doing that as part of the ritual/scam to hire h1bs. Keep looking and you'll find one. It's hard though because you're competing on unfair terms.

Just keep trying.

>>61754375
No problem. By the way, notice that you don't have to loop over the full length of the arrays when your data is sorted. Sorting itself is faster than a nested loop, so even when you you're not working with sorted input data, this solution will still be faster.

>>61754410
join military

>>61753700
I realised I can do this:

enum RETURN_VALUES {
    SUCCESS = 0;
    ERROR_A = 1;
    ERROR_B = 2;
    ERROR_C = 4;
};

int checkError(T* var, int flags){
    int return_value = SUCCESS;
    if(flags & ERROR_A && condition_A)
        return_value |= ERROR_A;
    if(flags & ERROR_B && condition_B)
        return_value |= ERROR_B;
    if(flags & ERROR_C && condition_C)
        return_value |= ERROR_C;
    return return_value;
}

int function1 (T *var){
    for(int errorCode = checkError(var, ERROR_A | ERROR_B); errorCode != SUCCESS; ){
        return errorCode;
    }
    //do stuff
    return SUCCESS;
}

int function2 (T *var){
    for(int errorCode = checkError(var, ERROR_A | ERROR_C); errorCode != SUCCESS; ){
        return errorCode;
    }
    //do stuff
    return SUCCESS;
}

int function3 (T *var){
    for(int errorCode = checkError(var, ERROR_B | ERROR_C); errorCode != SUCCESS; ){
        return errorCode;
    }
    //do stuff
    return SUCCESS;
}

int function4 (T *var){
    for(int errorCode = checkError(var, ERROR_A | ERROR_B | ERROR_C); errorCode != SUCCESS; ){
        return errorCode;
    }
    //do stuff
    return SUCCESS;
}

>>61754494
Or well, won't be worse of complexity-wise, anyway.

>>61754430
Should mention I got an MIS degree a few years ago but couldn't find work and had health issues.

>>61754495
I'm sure I'm too old for that option :(

>>61750753
That sounds fun thanks for the idea

>>61754547
Go into OCS.

Non meme response: Move to the Ukraine and start an outsource shop. People in countries like that can leverage the higher buying power of the American dollar against you to work for lower nominal wages but higher real wages.

Real non meme response: If you're a man you're fucked. Only women are allowed to be unemployed for long stretches of time.

Real real non meme response: Hope Trump kills H1B

Real real real non meme response: Try going to companies in person or sending an email rather than scouring through indeed For large companies emails are better. For small ones going in person is better.

Yeah, I'll just kms thank you very much.

Hope you have a wonderful day.

>>61754699
Don't you worry, I will. I'm a Haskell programmer.

>>61754699
Have you actually applied for the jobs or do you get scared off when you read the requirements?

>>61754547
http://www.suicide-do-not-kill-your-self.com/unemployment.html

>>61754320
It's not about compiling "before" or "after". By design, before the linker runs, these files don't know anything about each other (separate compilation model).
To fix this, you can #include a file containing my_def in both myClass and your main file so they both know whether it's defined

>>61754772
not him but the latter

>>61753963
https://zedshaw.com/archive/curing-pythons-neglect/

>>61754885
You shouldn't get too scared off. Don't be obnoxious and apply for a bunch of jobs at the same company because you'll get blacklisted, but for each company look for the most suitable job or two for you and apply.

Not apply for any jobs is worse than applying for jobs that you technically according to the words on the page don't match. So apply for the 5 most relevant jobs you can per day.

Try social engineering your way into jobs. Call the other people in your high school/college class. Go in person. Put that you're looking for a job on all your .README files on github.

Also you don't have to get a job for the degree you got. For interdisciplinary fields, a degree requirement just exists to show you don't eat paint. My father's got a degree in quality assurance engineering but he's worked in HR all his life.

>>61753992
There is no sane way to handle POSIX signals. There is literally NO WAY to do it that doesn't involve exceedingly complicated system calls and their arcane documentation. The only way that might somewhat approach something called "ok" is to globally turn off signals and then open a new execution thread that will do literally nothing but poll a signal file descriptor continuously and communicate events with the other interested threads.

Don't fall into the deep signal hole. It was a huge design mistake as far as unix is concerned.

aside from firecode and hackerrank what sites do people use for preparing for interview questions

File: a career in game development.jpg (100KB, 610x537px) Image search: [Google]

100KB, 610x537px

>wanting a job when you could just gamedev your way to billions of dollars like Notch

>>61754891
Yes this is perfectly on point. And isn't Zed Shaw the guy who wrote that really popular Python tutorial?

>>61754990
That one Cracking the Coding Interview book

>>61754993
Yeah just get extremely lucky with your one indie game even though the indie game market is saturated and the hype is all over. Have fun with that

>>61755006
congrats on getting the joke you fucking redditor

>>61755006

echo "fun at parties"

>wanting a job when you could just become a college prof and make 100k+ a year and get tenure

migrate
>>61755068
>>61755068
>>61755068
>>61755068

>>61754994
He wrote Mongrel, the Hard Way tutorials, Rails is a Ghetto and the Programming Motherfucker. Dude's the baddest character in the scene, bad enough to diss K&R's faggot C strings