forked from enlightenment/efl
parent
f843a2453a
commit
58965c4663
|
@ -41,7 +41,7 @@ mixture of these and more complex objects such as gradients, images etc. An
|
|||
application normally needs to hold all this data and when required, redraw
|
||||
the data do a window or pixmap if the data changes, or the window contents
|
||||
become damaged and need a redraw to restore them. To do this optimally the
|
||||
application would also need to figure out what exactly canged, which parts
|
||||
application would also need to figure out what exactly changed, which parts
|
||||
of the window or pixmap this data mapped to and then order the drawing so
|
||||
all objects are drawn in the right order, and only the part of the window
|
||||
that needs redrawing is drawn to to save excess processing. For a large
|
||||
|
@ -82,7 +82,7 @@ Infinite loop ...
|
|||
<p>
|
||||
In a naievely written program an event loop may have many re-renderings and
|
||||
re-drawings going on as things are modified, otherwise the application needs
|
||||
to retain state like evas does and ten handle draws in idle time too, which
|
||||
to retain state like evas does and then handle draws in idle time too, which
|
||||
means applications need to do the work Evas does themselves, as well as the
|
||||
optimizations already in place in Evas.
|
||||
<p>
|
||||
|
@ -98,6 +98,14 @@ many times over, without you having to know a single line of OpenGL. Evas's
|
|||
Software rotuines alone are many many many times faster than Mesa's software
|
||||
OpenGL rendering, as Imlib2's code is purpose written for 2D operations with
|
||||
Assembly optimizations for the core routines.
|
||||
<p>
|
||||
</blockquote>
|
||||
<p>
|
||||
<b>Canvas Principles in Detail</b>
|
||||
<p>
|
||||
<blockquote>
|
||||
<p>
|
||||
|
||||
<p>
|
||||
</blockquote>
|
||||
<p>
|
||||
|
|
Loading…
Reference in New Issue