Flash assignment - which one is least impossible?

Nima

I have a uni assignment in which I either:
a) Make a game that uses AI pathfinding.
This option would require that I make characters on an isometric or flat plain that have the ability to detect each other. One character is controlled by the player, the other is AI and must be able to detect and wend its way towards the player's avatar.

b) Make an isometric level design with the brief that it's designed with multiplayer in mind. I must present it beautifully and indicate where and what interactive/ mutable parts of the scenery are, and what they do (eg: depict a bridge and tell the marker that it collapses if too much weight is placed upon it, or the player can destroy it with dynamite - that sort of thing).

I'd prefer to do option a, but I'm not much of a programmer yet in actionscript. I can use flash reasonably well for design work, and program various bits and bobs at a basic level to do stuff like move up down, left or right, and the effects of gravity and acceleration in a very simple fashion. I have to admit that collision detection using programming, rather than the hit test function, really was too much for me at my current level, though how much of that was down to a bad explanation I can't say. I recently purchased a book called 'Flash 8 actionscript bible', which I'm hoping will help with either option.

Which option do you fine people think is least likely to make my hair go gray, considering my current inexpertise in actionscript (I want to learn but there's a deadline about 6 weeks away)?
Does anyone have any useful resources they've found for isometric grid programming in Flash 8?
Is Flashkit the best site I'll find, or am I missing something excellent?

Any advice very much appreciated!

Smasher

I don't really have flash experience, but if you go with option A the pathfinding algorithm widely considered to be the best for most cases is A*. You might be able to get away with something simpler (though A* isn't really complicated once you understand why it works) if the only obstacles are the computer AIs, but if you have static objects or walls or what not you'll probably want to dive in and do it right.

You can find a bunch of A* tutorials by googling "A* pathfinding".

Herschel

It seems A is a subset of B -- to have a working isometric tile game with AI pathfinding, you've got to design an isometric level, no?

Here's a real good set of articles that will take you through making a tile-based, and then isometric, game.

http://www.tonypa.pri.ee/tbw/index.html

GogoKodo

If there's nothing stating that it has to be done in real time then a turn based tile game would seem like it shouldn't be very far out based on the skills you described. You won't need to do any hit detection since everything is turn based and on tiles it should be pretty easy to do the AI.
If those descriptions you gave are the only description given for the assignment I would assume that they are talking about a turn based game, especially if they've only taught you the simple things you mentioned in your skillset.

Nima

Thanks very much for the advice, everyone.

Actually, they haven't taught us any of it, it's a pretty rubbish university.

Szechuanosaurus

Personally, I'd go with option A. There are plenty of tutorials on adapting A* pathfinding routines into flash via actionscript and it requires less 'building', by which I mean creating assets etc. You could essentially build it on a 2 dimensional field with plain squares as obstacles and circles as the player controlled character and the monster chasing them. From there you can rapidly build the AI routines and then if you have time, go back and flesh out the graphics into something more appealing and presentable.

That's definitely your best option if you are completely new to actionscript as it will give you the time to get the code right without having to spend hours and hours on graphics. With option B you have to worry about all sorts of complicated movement rules and getting presentable graphics that explicitly inform the player that they are working within an isometric field, not to mention developing some sort of overall concept for an actual game. With A you just need to worry about 1) How to give the player control over the movement of a movie object and 2) How to get the computer to control the movement of another movie object (via A* pathfinding).

Definitely option A, in fact.


I reckon I could get the nuts and bolts of A working in an afternoon and have something that could at least get a respectable pass, whereas with option B I'd probably take at least an afternoon just to come up with a rough concept and from there who knows how long to actually build the whole thing. With the time saved on option A, I could then go on to create an outstanding submission that earned me a merit via additional gameplay mechanics and better graphics. Ultimately, you might even have enough time to develop option A into option B.

So yeah, start with option A, see how it goes.

Nima

Szechuanosaurus wrote: »

Personally, I'd go with option A. There are plenty of tutorials on adapting A* pathfinding routines into flash via actionscript and it requires less 'building', by which I mean creating assets etc. You could essentially build it on a 2 dimensional field with plain squares as obstacles and circles as the player controlled character and the monster chasing them. From there you can rapidly build the AI routines and then if you have time, go back and flesh out the graphics into something more appealing and presentable.

.

Thanks very much, this was my feeling also. I'm just intimidated by the scripting, but option a gives me time to work on that. Thanks for clarifying that in my mind.