Any language or method is valid so long as you generate the correct answer (barring the method of looking up the answer elsewhere) but the focus most people take is to do it as fast as possible and in as few lines of code possible. There’s even a bunch of people solving the problems in assembly code. New mind please, mine just blew.

It’s a great test of your problem solving skills and a great way to learn new things about your chosen programming language. Do you know how to use BigIntegers or reverse a string in Java? I sure as hell didn’t, nor would I have normally looked it up. Learning!

If you’re at all interested (and not allergic to clicking links) you can find it here. It’s probably about time I showed off some source code.

A game designed for, and whose graphics can be fully represented by, a horizontal row of coloured pixels.

Sound simple? That was kind of the point. View it as a challenge, if you will; can you design a game that simple? I’m skipping over controls a bit, but they can be as simple as you need; d-pad and two buttons anyone?

Anyway, here’s my starting effort; touch a tile to cycle it and the surrounding tiles’ colour through red, yellow and blue.

The basic goal would be to get all tiles the same colour; initially I had this at 5 tiles, but since the tile colours are purely random it could result in unwinnable game. It was simpler just to make it six tiles for the prototype, but the game feels quite easy now. If I don’t solve the ‘unwinnable scenario with 5 tiles’ problem (or even if I do) I’ve had some ideas of things to add:

• All Red/Yellow/Blue: get all the tiles to a specified colour
• Match the pattern: manipulate your tiles to match a pre-determined pattern
• Speed mode: Get all the tiles the same colour, board resets. Repeat as many times as you can in a time limit.

So, how are you?

I won’t be long. Hopefully.

If you’re looking for something like this, moo.com is where you’ll get them. You can use your own designs on both sides or have one side with formatted text. Recommend it.

• My updated CV is available for download on the About Me page, or here.
• Added ‘Current Projects’ links on the right to make it easier to see what I’m up to just now
• New person on the Blogroll; Searra Dodds, artist and Lego appreciator.
• Added a new category to my work – Post-University work
• Added ‘Yet Another Misdemeanour’ to the aforementioned category

• I put up a new flowchart picture for the Knight’s Tour post. The new one is the correct size and actually makes sense. This is why Gordon should not try and construct flowcharts at 6am.
• I’ll be heading to Game In Scotland this coming weekend, which means two things; doing work for my reel and updating my CV. I’ve had some great feedback to go on, but my work is still cut out these next few days

So I’ll see you at Fat Sam’s this weekend if you’re looking for games industry employment or enjoy hearing Brian Bagalow talk a lot.

At first, I attempted a brute force approach; randomly select a move, starting again if the knight runs out of moves. This failed to actually produce a full path; so backtracking was implemented, allowing the knight to reverse when no moves are available and not all squares have been visited. Another technique used to ensure paths were found was to move the knight to locations from which the least number of consequent moves could be made. (i.e. move to whichever square provides the least options)

In order to learn the principles behind solving the knight’s tour, my first program gave a visual representation of solving the knight’s tour without recording or tallying total paths. However, the programming test specified that a method return the total number of paths for the grid specified in a console program. The basic logic of the program would go as follows:

Of course, given the number of possibilities that the problem produces (for an 8×8 grid, this surpasses millions) this is still theoretical. Hopefully I’ll be able to update this as I make some progress.