TEMP: start of deferred/on-demand processing
authorJon Nordby <jononor@gmail.com>
Thu, 10 Oct 2013 23:30:23 +0000 (01:30 +0200)
committerJon Nordby <jononor@gmail.com>
Fri, 11 Oct 2013 12:20:23 +0000 (14:20 +0200)
commit93066f4865eeb5af99c998289cdf517235d22f75
treeed22ea605bc6a91363864a63f7bdbba4f7437a86
parent8acd017a5301d5ec3c635b0ce98077cbe3723e38
TEMP: start of deferred/on-demand processing

Idea is to store which stroke_to() dab operations correspond to in the op queue,
and instead of processing everything in queue at end_atomic(), we just return the dirty bbox.
The application requests a redraw for this bbox and in the draw_cb() calls process()
It can also limit the time process() takes by specifying to compute only up until a certain
stroke_to() operation, and use the return value to request a new redraw for the areas which
are still dirty.
brushlib/fifo.c
brushlib/fifo.h
brushlib/mypaint-brush.c
brushlib/mypaint-surface.c
brushlib/mypaint-surface.h
brushlib/mypaint-tiled-surface.c
brushlib/mypaint-tiled-surface.h
brushlib/operationqueue.c
brushlib/operationqueue.h
lib/tiledsurface.hpp
tests/test_mypaintlib.py