| 
GEOS SDK TechDocs
| 
| 
1 Geometry Manager Overview
| 
1.2 How Geometry Is Managed
The geometry manager, used effectively, can take nearly all the work out of displaying, sizing, and positioning objects on the screen. Many of the tasks you normally would have to do to create even a simple dialog box are taken care of automatically by the geometry manager and the specific UI library. Among the functions the geometry manager performs are the following:
-
Window Positioning and Sizing
The geometry manager interacts directly with the UI to determine where windows should be placed and what their dimensions should be. You have some control over this with hints, messages, and object instance data, and often you will have to choose a hint to determine the window position or size.
-
Object Sizing
The size of an object can be determined by the combined sizes of its children, by the limits of its parent, or specifically by other constraints on the object. The geometry manager allows all three types of sizing behavior. The first two are implemented automatically in geometry updates; the third depends on how the application wants to manage the "other constraints."
-
Object Positioning
The position of an object can be determined either by the object or by the geometry manager. The geometry manager can position objects based on several criteria including the spacing of siblings in a composite, the size and/or location of the parent composite, and the justification hints applied to the object. Visible objects can control very closely how and where they are positioned; generic objects can determine how positioning occurs with various hints.
-
Automatic Geometry Updating
Typically, when a particular object in a tree is resized, repositioned, added, or removed, the object mechanisms will mark the geometry of both that object and its parent invalid. This will invoke the geometry manager, which will traverse the object tree, determine which objects are affected by the potential geometry change, and then update the size and position of all affected objects up to the top of the tree. This entire process is automatic for generic objects; for visual objects, you may or may not have to mark the object invalid yourself, depending on the depth of custom geometry control you use.
With geometry functions covered by the geometry manager and drawing functions covered by the specific UI library, you have little else to do but specify the structure of your object tree and the hints each object should have. The system does all the dirty work of drawing and managing the objects on the screen, leaving you free to work on your application rather than on its graphical screen representation.
| GEOS SDK TechDocs
|
| 1 Geometry Manager Overview
| 1.2 How Geometry Is Managed