Sup Forums Makes a List, Part 4: List of Lists of Lists

We're nearing the end of this little endeavor, so that's good! At the request of several anons, I need 100 challenges, and we have about 110...

Which ones should I take off?

Other urls found in this thread:

youtube.com/watch?v=J691aLfkWP0
github.com/freecode/dominos-cli
youtube.com/watch?v=qQNnJQIPHV4
jsfiddle.net/46xxsvkg/
fortytoo.github.io/
youtube.com/watch?v=ka5KpaKDGF0
en.wikipedia.org/wiki/Sudoku_solving_algorithms#Constraint_programming
twitter.com/SFWRedditImages

my idea

kek

I love that webm

roll

actually...
youtube.com/watch?v=J691aLfkWP0

that would be a cool programming challenge..

That's a goddamn genius challenge. Adding it know... What difficulty?

>Which ones should I take off?

Sort them by category (games, AI, algorithmic, everyday software, compiler/interpreter/debugger, emulators...)
Then sort them by difficulty.

If you have to remove some programs of the list, see which category and difficulty have too many projects.
And it will be easier to find out what to remove.


And something I'm wondering, what is better : having a random list that you can roll or an organized list so you can pick something you actually want to do ?

what would be difficult is to find the API and how to use it. The program might not be very difficult for what I can read here
> github.com/freecode/dominos-cli

Organized List all the way... But yeah, I'm in the process of categorizing them!

Here's the progress of the sorting so far.

>which one should i take off
none, i'd say the last 10 challenges should be exclusively for trips get, i.e. change 102 for 222 and so on

looking good

Part 1 of the sorted list

Part 2!

"Make a program that produces this .png with a list of programming projects"

That's... Actually not that bad of an idea

Better version of the sorted list

Not that I've done something exactly alike but I think it should be easy with the right library

Cluster computing something.

Hm... I'm not sure if I should include it on the basis that you'd need more than one computer (and not everyone has access to that)

Virtualization.

As in Virtual Machines, right?

Added some challenges from v2.0

yeah, and how many of you have individually coded any of the "hard" or "fuck you" challenges?

Going off of the list in ,
>written ray tracer for a class in uni
>written basic flying camera stuff in open gl for same class (sorta like a first-person engine)
>I have an esoteric langague I've designed, implemented a reference implementation (need to get around to doing some optimizations), and proven turing completeness of for fun
>shitty half-written text editor to see what writing one functionally would look like
>shitty unoptimized turing machine sim

Not too bad desu

Goddamn. I'm on that weird level of skill where the easy challenges are too easy, but the medium challenges seem a little too difficult. I guess if it isn't hard, you're not advancing. The only medium challenge that sticks out as maybe being doable for me is Dijisktra's Algorithm.


Also, what the hell is "Pipe Dreams"?

It's kind of weird explaining it, but here's a video:
youtube.com/watch?v=qQNnJQIPHV4

I thought it was pic

That picture just sent a flood of pre-2005 memories rushing through my mind... I would add that as a challenge, but that could fall under "64KB Demo"

Hoping for something easy - I'm a freshman... rollan.

hmmm, would the challenge then be to implement my own version of Pipe Dream?

Rolling. If I get Hard or worse move to closest medium

Roll

That's fine! As long as it works!

I added some neat things to my game of life
jsfiddle.net/46xxsvkg/

By I was wondering if it's possible to go "back" in generation, something like a ctrl+Z.

Is there some reverse algorithm to go back or do I have to save previous generation stat ?

Holy fuck this is amazing... Things like these are why I started the 4.0 list in the first place.

Theoretically speaking, couldn't you make some sort struct where it keeps a history of the states of each square?

Yeah I would just save the arrays contenting previous states each time, but the would be very heavy
I'm think it's possible to have a "reverse rule" who would play the game backward

Doing the inverse Game Of Life would be loads more efficient... I recommend you do that

Btw moved it to github, jsfiddle isn't working well fortytoo.github.io/

>fortytoo.github.io/
Lookin' good!

>chatbot
>easy
I think I'm misunderstanding that project.

just do a simple AI that understands human language and can talk back to you lmao

thats like 10 lines of ASM you shitter

>Is there some reverse algorithm to go back
No - say you have a 2x2 grid set the following two ways:
X-
-X

and
-X
X-

both result in an empty grid

therefore there is no injective function reversing game of life

So we're having the final version ITT? I don't know how many versions I've saved already just in case I miss it.

109 should be easy
108 should be easy
107 should be medium+
79 should be medium
65 should be easy if 99 is easy

difficulties are fucked up all over the place

Yup I just read about it, it's crazy how non-deterministic GoL is and how it resemble entropy

>40: Sudoku solver using A* algorithm
What does this even mean? A* is a shortest path algorithm. It doesn't have any bearing on sukodu puzzles.

GoL is fully deterministic

A* is an informed (uses a heuristic specialized for the problem) search algorithm that you can use to find a shortest path in a graph.
Search algorithms can be applied a range of problems such as constraint satisfaction problems.

(Me)
I advise you to listen to this grand lecture: youtube.com/watch?v=ka5KpaKDGF0

>Search algorithms can be applied a range of problems such as constraint satisfaction problems.
What, by searching the implicit graph of all exponentially many partially-filled sudokus for a given puzzle? That must be stupendously inefficient.

Well no.
It's not that bad.
While CSPs usually cannot be solved in polynomial time, you don't need to check every partially filled sudoku.
CSPs (and sudoku is a CSP) have the nice property that it doesn't matter which variable you assign first.
You can pick the one which are forced moves because only one possibility exists first (Minimum Remaining Values) and then the value that rules out the fewest options within other variables (Least Constraining Value).

Using these two heuristics sudokus can already be solved quite fast.

Yeah, but that's not an A* solution. I was discussing the use of A* for solving sudokus.

I suppose you could forcibly fit all that special case handling into a cleverly encoded A* heuristic, but then you are clearly using a solution in search of a problem. Yeah, you could do it, but it's retarded and still inefficient.

>Yeah, but that's not an A* solution. I was discussing the use of A* for solving sudokus.
Solving CSPs can involve using an informed search algorithm one of which the A* algorithm is.

>it's retarded and still inefficient
I'm not bullshitting you here, this is entry level AI.
en.wikipedia.org/wiki/Sudoku_solving_algorithms#Constraint_programming
What would be your way of solving sudoku?

>Solving CSPs can involve using an informed search algorithm one of which the A* algorithm is.
Yes, they can involve informed search algorithms, but that doesn't mean it should use A*. That's like saying "I need a tool to put this nail into this piece of wood; a wrench is a tool; so I'll use a wrench". A* is an informed search algorithm, but it's not a CSP algorithm.

>What would be your way of solving sudoku?
Use your favorite informed search CSP algorithm. Which A* isn't.

Yeah you're right, I was talking about reverse-GoL, which is not deterministic ?

>Use your favorite informed search CSP algorithm.
Give an example of an "informed search CSP algorithm"

Yeah I guess you could call reverse-GoL non-deterministic.

Rolling

Nice I'm not home atm so first person to reply to this picks programming language

Heuristic-based (that is, "informed search") backtracking with forced moves is one, for example.

IDA*, SMA*?

I'm going to do all of these using nothing but the standard python 3.2 library.
Wish me luck :^)

What language would you guys recommend to do those easy games (snake, etc)

I know how to do the logic but I've never done GUI programs before.

That's literally what I described. A* is just the search algorithm in use which you can join with MRV and LCV.

A* will probably work just as fine as IDA*, i.e. I don't think memory will become a problem.
SMA* is neither guaranteed to be complete nor optimal and slower compared to A* if the memory bound is low.
I encourage you to try it out, though.
They will probably both work all right, but I don't see what qualifies them as a "CSP algorithm" but not A*.

PL/I