Hey Veeky Forums...

Hey Veeky Forums , /g/ is making a list of programming challenges if you have any suggestion post in this this thread or in this one

Other urls found in this thread:

projecteuler.net
blog.codinghorror.com/why-cant-programmers-program/
twitter.com/NSFWRedditImage

Make a program that takes a grey scale square image of a number, and output what digit it is with more than 75% accuracy

Make a jpg to png converter.
This is way harder than a VM.

a program that can find OP's cock (NP hard problem)

Quite a general on this:
>Create a lattice QCD alogrthm that does better than [math] \tau \propto a^{-2} [/math] where [math] a [/math] is the Lattice spacing.

Recommended difficulty level:
>Difficult (at least)

If you only search in males’ asses then it becomes O(n).

Half of these are cool, the other half is worthless interface/network shit.

Also, a VM with a program can print strings is trivially easy. I did that in less than I day while I was 15 on a graphing calculator while crammed into a car with 4 other people. (It was dammed slow though) Just add ~20 lines of code or so to the RPN calculator, there's a reason stack machines are popular.

How many pixels per image will there be?

The observation that comes to mind is that both of your purple challenges are in fact easier than most of the red ones.

Fuck OP I want 12 so badly
>feelsbadman

projecteuler.net

Monster Rancher, there is a GBA version if you want to play on mobile. Otherwise i would go for the PSX version. Them music.

>How many pixels per image will there be?
4

>Basic VM
>Bootloader
>Fuck you tier

There are lots of computational algebraic geometry things, like finding the number of rational points on an elliptic curve, or the number of points of some variety over a finite field. These could be pretty easy or incredibly difficult, depending on what exactly you ask - you could find the number of points of a variety over a finite field by checking, but that would be slow.
There are obviously lots of tricky physics simulations, like lattice QFT or numerical relativity.
Something at the intersection of all of this might be "generate a list of Calabi-Yau 3-folds using the Batyrev-Borisov construction, and their Hodge diamonds". It has physic-y applications , and the construction itself comes from physics, but it is really a computational AG problem.
It is not easy.

>Sudoku/n-Puzzle solver using A* algorithm.

Can someone please explain why A* is necessary? I can see how you would model it as a graph problem and how to solve it using BFS on an implicit graph of all puzzle states (since explicit would be retarded to do for sudoku, and unreasonable for large n on n-puzzles, although for n-puzzles you'd have to worry about what you've visited so you'd need a good mapping of states to something smaller like an int I guess) (or backtracking, which I guess would be a DFS on the implicit graph), but why the hell would you need A*? What the fuck would the edge weights be?

>fizzbuzz
>challenge

Seriously?

Hence why it's in the easiest category. The purpose of the challenge thing is for people who are learning to code. It's a good problem for someone who is just starting out.

Is Verlet integration better than leapfrog integration for calculating orbital trajectories?

Well, that complicates matters.

blog.codinghorror.com/why-cant-programmers-program/

Fuck that shit is amazing. It really tells the whole story of the CS degree and software engineering education in general.

However, I don't agree with judging them by them being able to write for loops in the languages they list in their CV and that is for personal reasons.

I am a C/C++ guy but in the job I do my language distribution is:
50% visual basic
45% Java
5% C++

From that you'd think that I am a genius at visual basic but honestly I do not give a fuck about the specifics of that language. I GOOGLE SO MUCH SHIT.

Let me tell you, I know that for each loops exist but I do not know the syntax. I've had to write them when dealing with specific components but I just google the syntax and then copy and paste it and change the variables to fit my needs.

That said, I could do fizzbuzz in vb using regular loops which I think I've memorized for now. Plus I could do them no sweat on C or C++. Java too, C style languages are just objectively better when it comes to memorizing structure.

Knowing how to do for loops in languages does not mean you are a good programmer. Hell, I've done some crazy fucking code-contortionist level shit in visual basic and yet I do not know how to write for each loops.

I do. The inability to write something that basic means they really don't know how to program or understand even the most basic ideas of computer science/programming. Maybe they have other skills or knowledge but they are not programmers.

That's like saying a musician is shit because he didn't memorize the music sheet.

How do you have a job again?

I can't even believe you're arguing this? What the fuck!?
No this is like saying a musician is shit because they can't play an instrument, which is a perfectly reasonable thing to do!
Getting a job is a social exercise and has almost nothing to do with skill (although the ability to convince people you have it helps, either by acting or with a legitimate demonstration of knowledge)

SKI combinator evaluator.

A* isn't necessary or even better I think. but makes it a little more interesting.

Way I would do it is say the heuristic cost would be 1 for each in occupied space, while the actual cost is the number of occupied spaces plus the number of spaces who are in conflict because of the potential move

Not him, but I list like 8 languages on my resume.
I couldn't tell you off the top of my head what the syntax for for-loops were in Java, but I could tell you about them in four other languages.

Most coding is done in front of a computer with internet. I can see the argument both ways, that if you can't code without the Internet, you're nothing but a glorified technician cutting and pasting together stuff you don't really understand.
On the other hand, what matters in the real world is the final result. They don't tell you have to do Gauss-Jordan elimination by hand, why should you restrict yourself in CS?

If you don't know how to do for loops you must be programming something fucking mind numbing, also I refuse to believe you're a good coder if you don't know the for loop syntax for the language you code in

I haven't used Java in 7 years, that doesn't mean I couldn't just pick it up again if I was refreshed on the syntax.

Think I linked the wrong people was talking about the Visual Basic cunt

>Least Squares
>medium
Maybe if you can have more than one functional basis in the program and add weights to the points.

Your argument applies to him to though.

You're arguing that it's reasonable to ask programmers what the for-loop syntax of any language they list on their CV is and they should know it off the top of their head precisely.

find SVD and ure good

...

That's not at all what I said. I said if you program in a language, I'll make this more clear to say that if you use a language regularly ( > a couple of times a month) you should fucking know the syntax for a for loop, if not you're a sad excuse for a programmer

>RREF
>easy
This is something that anyone with even minimal programming skills should be able to do.

Yea I don't think they're meant to be difficult, just challenges to keep sharp on some basic programming skills, maybe work a data structure or two

If you can understand multi-dimensional arrays and for() loops, you can code an RREF calculator.

For loop syntax is not complex at all, if you don't know it then you haven't used enough to be very familiar with them or you lack the fundamental ability to understand them (I personally doubt the second one is actually possible but I can't be sure).
This is bad because often you have to read and contribute to other people's code which requires some understanding.