Example 2:
 This example demonstrates how to insert bitmaps in the world and how to
 use these bitmaps to build sprite frame descriptors. It also demonstrates
 the pixel precise collision detection routine of BETATRON.

 When working with sprites you must remember that a bitmap cannot be assigned
 to a sprite. Only a frame descriptor can be assigned to a sprite. So ,
 generally, bitmaps cannot be used directly which means that every time
 you insert some bitmaps in the world you also have to create some frame
 descriptors. You assign a frame descriptor to a sprite by setting values in
 TOsprite::framenow and/or TOsprite::framebase.

 Notice that the sprite representing the 2nd sign (TOsign2) is a descendant of
 TOsprite. It carries one new field: a pointer to the other sign. 
 This pointer is used for the collision detection check inside sign2action()
 Of course, there is way to check an object for collision against an entire set 
 of sprites by using a 'clustering grid'. This is shown in the next example.
 
 The TOworld::collision() routine works in 2 different modes. 
 In the first mode the collision masks of the frame descriptors assigned to the
 sprites are used for pixel precise collision detection. If both sprites have
 collision masks then this mode is selected automatically.
 In case that at least one sprite does not have collision mask then 
 TOworld::collision() selects the second mode which uses the bounding boxes of
 the sprites for the collision detection. In the last case the collision 
 detection is somehow poor.

 To see TOworld::collision() working in the 'poor' mode comment out the call:
    myworld.makeframecolidmask(1);
 or pass a 0 as the third argument of the call:
    myworld.make1bitmapdescr(0,0,1);
 and then recompile the example program.