Snapping to a grid.

Here's a quick explanation. Think of the intersecting points where each grid is devided, that is where you want your points to snap to right (of course I know you already know that :))? Now think of that point as being at the center of a box the same size as your grid resolution. When your cursor, or whatever enters that box, set the point to the intersection of the lines at that grid location.

I hope that was clear enough. Post back if you have any questions.

First of all I think you'll need to store the position of the mouse where it was clicked and where it is now. Then you'll perform the operation on the object depending on those two points. Let's say we're translating: you'll add a vector (equal to current mouse minus initial mouse positions) to the object's position. I would suggest you don't actually change the position but use a temporary position (for unexpected errors caused by the user).

After that's done you'll do the snapping. You can snap to vertical or horizontal grid line and grid point. For the horizontal lines: let ty be your temporary y coord and ly be the line y coord. If abs(ty-ly) < treshold then you set ty=ly; Same for vertical lines only use the x coordinates. For points check both y and x coordinates.

To determine the grid lines where you could potentially snap just take your temporary x and y coordinates and find the closest multiple of the grid size.