kangax
|
2ade9dc0f8
|
More optimizations: setCoords doesn't need to be invoked on an object before it's added to canvas. This makes most sense when creating path groups, since each of the paths in a path group doesn't really need to have coords. Coords are used to determine object boundaries when moving/scaling/rotating objects, and paths in pathgroup never participate in it (instead, pathgroup does).
|
2011-02-09 17:30:35 -05:00 |
|
kangax
|
45c9a333a8
|
Few optimizations in _parsePath implementation.
|
2011-02-09 16:35:46 -05:00 |
|
kangax
|
d42ba6ec02
|
Add links to benchmarks in readme; modify simple shapes benchmark to be able to chose number of shapes.
|
2011-02-09 02:27:25 -05:00 |
|
kangax
|
0a1c52408d
|
Add breakdown of parsing vs. rendering to complex_shape test.
|
2011-02-09 01:09:30 -05:00 |
|
kangax
|
83d1d7cecf
|
Uncomment raphael test in complex_shape_2.
|
2011-02-09 01:04:31 -05:00 |
|
kangax
|
ca11620451
|
Replace map with for loop for perf. reasons (as this method is likely to be called thousands of times).
|
2011-02-09 01:03:04 -05:00 |
|
kangax
|
afb841e085
|
Add another complex shape (~9K paths) to the raphael vs. fabric perf suite.
|
2011-02-09 00:41:27 -05:00 |
|
kangax
|
ba4486e5be
|
Use Array.prototype.slice to convert nodelist to an array (slice is usually faster; it makes sense in this context where nodelists (i.e. shapes) could consist of thousands of elements).
|
2011-02-09 00:14:12 -05:00 |
|
kangax
|
4d86890b68
|
Few tweaks to raphael vs fabric tests.
|
2011-02-06 04:28:04 -05:00 |
|
kangax
|
7e395b9a85
|
Few tests comparing performance of fabric vs Raphael.
|
2011-02-06 04:10:08 -05:00 |
|