1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
|
Design of an immediate graphical user interface api with managed layouts
Terms:
1. Canvas - a drawing area with a managed layout
User stories
1. I would like to call UI-rendering functions without specifying dimensions and locations
2. I would like to call custom UI-rendering functions while specifying dimensions and locations
Story 2 needs a way to get a proper starting location or a bounding box If 1 is in place.
Consider the single method Layout interface:
type Layout interface {
// Advances the layout and returns a starting point
// for an element based on the last ending point.
Next(Point) Point
}
Story 1 necessitates that library-aware UI-rendering functions acquire their starting locations
on their own. This allows for the following two approaches (more are possible of course)
* Store the ending point of the last element. New calls can get it and pass it to the layout
+ does not advance the layout unless required, saving cpu
* Store the next starting point by calling Next() at the end of the element drawing routine.
The next drawing can continue from it.
- advancing the layout costs cpu
- the user might want an out of layer render
|
|
>
>
|
>
>
>
>
>
>
>
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
|
Design of an immediate graphical user interface api with managed layouts
Terms:
1. Canvas - a drawing area
User stories
1. I would like to call UI-rendering functions without specifying dimensions and locations
2. I would like to call custom UI-rendering functions while specifying dimensions and locations
/----------------------------------
Story 2 needs a way to get a proper starting location or a bounding box If 1 is in place.
Consider the single method Layout interface:
type Layout interface {
// Advances the layout and returns a starting point
// for an element based on the last ending point.
Next(last Point) Point
}
Story 1 necessitates that library-aware UI-rendering functions acquire their starting locations
on their own. This allows for the following two approaches (more are possible of course)
* Store the ending point of the last element. New calls can get it and pass it to the layout
+ does not advance the layout unless required, saving cpu
+ does not imply a dependency on the layout, the UI drawing needs just a starting point
* Store the next starting point by calling Next() at the end of the element drawing routine.
The next drawing can continue from it.
- advancing the layout costs cpu
- the user might want an out of layer render
- puts a dependency on the layout
The first strategy is definatelly more sound and neatly allows for a layout-aware tracker
to handle the locations.
\----------------------------------
3. I would like to enhance drawing functions with additional behavior (e.g. layout trackers, styling, etc)
|