| commit | c4dd1c0ec8d8ceecf29bcccb5037edc7df95c55f | |
| author | Julian Lamb <thepurlieu@gmail.com> | |
| Wed, 15 Oct 2008 22:02:46 +0000 (15 22:02 +0000) | ||
| committer | Julian Lamb <thepurlieu@gmail.com> | |
| Wed, 15 Oct 2008 22:02:46 +0000 (15 22:02 +0000) | ||
| tree | 57117c2e1b9b6d77b73c1d168a80f37b7ac8553d | treesnapshot (tar.gz zip) |
| parent | 5faadda7b57a1cf4d9c77570a8da4942950dd5dc | commitdiff |
* pre-calculate a bounding box for every object during file parsing, and use the box dimensions to determine if that object will even be rendered inside the visible window. This approach allows for slightly better speed than the previous version, since it now includes polygons in the optimizations. My test boards show about a 25% speed improvement with this new method. The autoscale feature is somewhat broken right now...excuse the mess while I finish debugging things.
* also, apply the optimizations to "FAST" render mode as well...looks to be about a 3x speedup when zoomed in to a board.
* also, apply the optimizations to "FAST" render mode as well...looks to be about a 3x speedup when zoomed in to a board.