Wesley, Useless Hacker is right: for your style of isometric engine you don''t need anything as complicated and fiddly as mousemaps. You can simply convert between screen and tile co-ordinates using two simple formulae, already provided by Useless Hacker above - although you''d have to modify them a bit to suit your exact system. Which I''ll do for you now actually:

Given the size of your tiles (64x32) and the way you order them (you have the origin as the left-most point), to get the tile co-ordinates from the screen co-ordinates you use (in C):

tx = ((sx - mapx) / 64) - ((sy - mapy) / 32);
ty = ((sx - mapx) / 64) + ((sy - mapy) / 32);

where:
(tx,ty) are the tile co-ordinates
(sx,sy) are the co-ordinates of the mouse on the screen
(mapx, mapy) are the co-ordinates of the map''s origin relative to the upper-left corner of the screen.

To understand what (mapx, mapy) represent, look at the image you gave just above (which I''ll try to insert here, but I don''t really know how to use this forum, so it may not appear):



If you assume the white rectangle is the screen, then the origin of the green tilemap (which is halfway down the left side of the rectangle) would be at (0, 80). Therefore mapx = 0 and mapy = 80.

[And if the player walked down the screen, the tilemap would scroll upwards, mapy would decrease. If the player walked to the right, the tilemap would scroll to the left, and mapx would decrease. And so on.]

So if the mouse was clicked at some random point, say (100, 40) you would calculate the tile co-ordinates thusly:

tx = ((sx - mapx) / 64) - ((sy - mapy) / 32);
ty = ((sx - mapx) / 64) + ((sy - mapy) / 32);

so:
tx = ((100 - 0) / 64) - ((40 - 80) / 32);
ty = ((100 - 0) / 64) + ((40 - 80) / 32);

tx = 1.5625 - (-1.25) = 2.8125;
ty = 1.5625 + (-1.25) = 0.3125;

The tile co-ordinates are precise; truncate them to get the index of the tile: (2,0) in this example. If you look at the map and roughly trace along 100 pixels, and down 40 pixels, you will see you end up on tile (2,0).

The maths involved here is incredibly simple: plonk your finger on the left-most point of some tile on the screen then move it (your finger) to the right 64 pixels (until you''re at the tile''s right-most point). You increased the screen''s X co-ordinate by 64, and the tile''s X and Y co-ordinates increased by one each. So now you know how the screen''s X co-ordinate affects the tile co-ordinates. Now do the same action from the tile''s top point to its bottom point. You increased the screen''s Y co-ordinate by 32 pixels, and increased the tile''s Y co-ordinate by one and decreased the tile''s X co-ordinate by one. Combine these and you end up with the formulae above.

The other way to think of it, of course, is to say that at some point you''ll know that the player is standing at some given tile co-ordinates [eg. (3.4, 2.25)] and you want to draw the image at the corresponding screen location. So you must have some way of converting from tile co-ordinates to screen co-ordinates - simply reverse the process (which gives the formulae above).

Sorry to waffle on like this; the process of converting between tile and screen co-ordinates for a simple (no elevations, etc.) isometric tilemap is so easy to do but so long-winded to explain in general.

The above DOES work for your type of tiles (assuming my assumptions above were correct) - if I''ve been unclear about anything in this too-long post, just ask.

-hotstone