Just a quick question.

What vb6 code do I need to set the mouse coordinate to a custom point?

Also, while I'm here, is there code to enable/disable the mouse cursor visibility, other than using a null cursor graphic?

thanks

Here is one I remember from waaayyy back in the day:

http://pscode.com/vb/scripts/ShowCode.a ... 5&lngWId=1

_________________
NetGore - Open source online RPG engine

From Spodi's link, the API declaration that matters to your setting the mouse X/Y:



And here's a wrapper that uses that and allows for you to pass a window's handle in case you want to give an X/Y that is relative to that window:
http://www.devx.com/vb2themax/Tip/19125

And here's info on hiding/showing the mouse pointer:
http://www.devx.com/vb2themax/Tip/18490

HTH,
-Bryk

_________________
www.mwgames.com ... Dicey Curves, Space Mission, Jump Gate, Gem Raider, DareBase, Castle Danger, Keeps & Moats Chess

Thanks both. With your help I've got it working now.

What I was doing was looking into my idea for a real-time worms game. I had to dig out my old copy of the game to remember how the pointer pushes around the map, and discovered that as the pointer nears the edges, the game slides it back at the same speed that it scrolls the map, so it just feels like a really wide virtual screen... Hence the need for the mouse API.

At the moment I've got a 4096x1536 pixel game world. I've created some projectiles and got them to fly about and crash into the terrain and explode out a circle of it. Then I started on collision detection and had projectiles bouncing... which wasn't the best bouncing I've seen. (It's not obvious how to find normal vectors from bitmapped terrain.)

I've had some trouble with collision detection though. Each delta-time step the projectiles are moving 20 pixels or so, and sometimes that step takes them inside the terrain, (or could miss a tiny lump of terrain if it skips over it in one time step.)

So I've changed this and introduced a pixel by pixel stepping within each delta-time step. But with 100 projectiles, delta time increases, causing more pixels being traced next time step, causing delta time to increase again, etc. And the program steadily grinds to a halt, and numbers explode because of the large delta-time, causing it to crash.

I could buffer the terrain to an array instead of sampling pixels from the bitmap for the collision detection. I expect that will give me more speed. But if that isn't enough, I'm not sure how I'll fix this.

Working with an array should be much faster than reading pixel color values every loop. {It would also allow you to change up your on-screen background color (or use a background picture) and only track the terrain bits in the array. But that might not be a value-add for what your doing.}

That will definitely speed the "nearing" portion of your collision routine

I know you like to do real-life math works wrt to physics in your games, but in this case you may need to find a good-enough efficient approach. Once you've picked the impact pixel, can you do a quick sample in a 3x3 (or 5x5) grid centered on that pixel to determine the slow of the nearby terrain and do flat-surface collision responses based on that sampled normal?

-Bryk

_________________
www.mwgames.com ... Dicey Curves, Space Mission, Jump Gate, Gem Raider, DareBase, Castle Danger, Keeps & Moats Chess

how about once you get a collision with the terrain, do a pixel by pixel check back along the path of the projectile until you get to the edge of the terrain, and explode it there.

_________________
The path to hell is paved with clever code.

That's a good idea Moglor But I'm not sure it would work for collisions with small isolated lumps of terrain. Would it just step over them, and not collide?

btw, I haven't updated this game yet. I've got about 20 projects going on at the moment. Tonight I'm finishing the levels for ISIS, and altering some of the earlier ones around a little

you could do a pre-calculation of where the missile will explode before you launch it. you would get a more precise ending destination because you wouldnt be skipping large chunks. and then after the missile is launched, when it is >= the predetermined destination, explode it at the destination. the user will never know.

i'm thinking Bryk is right about making an array to improve speed.

_________________
The path to hell is paved with clever code.