Crossroads II Remake?

[valis]

Gents,

I'm looking at doing a remake of sorts of Crossroads II: Pandemonium, a type-in game for the Commodore 64 that appeared in the pages of Compute.  It's basically a much more frenetic and complex version of Wizard of Wor, and you can see it here:

http://www.youtube.com/watch?v=6_VMgQ5NQqM

Before I get too far into this I'm wondering how I should do collision and steering.  I was thinking of having a system where the corridors are two creatures wide, so that there is some slop when entering tunnels and a tiny bit of dodging room within the tunnel.  I am also going to use a "Geometry Wars" sort of control scheme rather than the Atari 2600 "shoot where you're going, 8 directions" method.

Any thoughts on this?  I have collision detection routines and tilemap scrolling sorted.

[Shockwave]

It's a cool game to remake, I haven't played it myself but the video looks great.

It should be rather like making a Pacman game to make this and as long as you go completely over the top with explosions and particles you'll be on to a winner.

[valis]

I don't think I phrased my question very well.

What I am wondering is how, in Pacman, does the maze collision detection work?   It's thornier than one would think.  What I mean is, let's say you are going down a tunnel with a hallway coming off of it, and you want to turn into the hallway.  How is this handled at a game level?

Guess I can just answer my own question with some interface tests.  The AI is another kettle of fish... programming is full of these things that seem simple on the surface but contain hidden complexities.

I agree that it will be a fun game.  I am going to add monster generators like Gauntlet, and procedural mazes.

[Shockwave]

Ah I see, well as your maze is going to be made out of tiles then;

XP=(int(xpos/tilewidth))*tilewidth
YP=(int(ypos/tileheight))*tileheight

Would return the position in the maze of your character as an X+Y position that could be referenced against your tile map array, this way you can see what tile you are on and guessing that the tiles will be in a 2 dimensional array it will be rather easy to sample the surrounding tiles too to see if you can move onto them.

I did something like this a long time ago when I wrote a version of Pacman and it worked well.

If the above doesn't make sense, I'll type a better explanation tomorrow when I've got my glasses on and can see a bit better.

[valis]

Well, I have the collision lookup itself sorted.

My bigger concern has to do with this: given eight-directional movement (WASD) how do you move the player object into a tunnel which is at 90" to the current tunnel?  Or, rather, allow him to do so?

I am starting to think that this may be more simple than I think it is and that I am overthinking the problem so I need to actually sit down and start coding a test.  I have scrolling tilemap and collision lookup and I am thinking the rest will come.  after all, they did it in the 1980s...

AI will also be a somewhat interesting problem....

I'm used to programming Asteroids-like arena shooters where there are no tilemaps.  Tilemap collisions have proven to be surprisingly tricky in many ways.

Edit: It appears that the solution to my problem involves making the detection boxes for the sprites somewhat smaller than the max size of the tiles.  Then the whole problem is avoided.  Duh.

Also, Shockwave: I'm a legitimate poster and not a bot.  I'm very interested in demoscene coding and retro games, as well as Blitz and numerous other languages.  Is there any way you can take me off the captcha list so I don't have to tell it what continent Zimbabwe is in every time I want to post?

[Shockwave]

Also, Shockwave: I'm a legitimate poster and not a bot.  I'm very interested in demoscene coding and retro games, as well as Blitz and numerous other languages.  Is there any way you can take me off the captcha list so I don't have to tell it what continent Zimbabwe is in every time I want to post?

I need to look at this in the near future, you'll come off the list automatically when you have 5 posts though.  I know it's over the top but sadly some people seem determined to register dozens and dozens of fake accounts so they can backlink scam thier shit products and the questions are there to stop the human spammers, not the bots, the automated ones have all been defeated by other measures I've taken.

If I knew where any of these people actually lived, maybe it would be fun to go to thier house or place of business and put posters advertising a load of useless services on their premises to see how they like it.

[valis]

Oh, it's not really that over the top considering the lengths these people go to.  It probably goes some distance toward keeping idiots out as well.  I'm involved with a forum and in the last few days we've had THREE HUNDRED spam accounts registered, all from Russia.  The appearance is that a large spam network there has dropped email spam entirely and decided to retool for forum spamming.

We'll see how all that works out for them...

[Shockwave]

I'm involved with a forum and in the last few days we've had THREE HUNDRED spam accounts registered, all from Russia.s out for them...

If it helps you any I would be happy to share the technique of what I did to stop the automated bots with you, though I won't post it in public for obvious reasons.  Since I made these changes, no automated bots have managed to register here. Only humans.

[valis]

Oh, I am sure I could figure out some sort of turing test.  Problem is, our users are so stupid that no real humans would be able to register either