[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Gnash-commit] gnash ChangeLog server/parser/shape_character_d...

From: strk
Subject: Re: [Gnash-commit] gnash ChangeLog server/parser/shape_character_d...
Date: Mon, 12 Nov 2007 21:09:26 +0100

On Mon, Nov 12, 2007 at 07:24:29PM +0000, Udo Giacomozzi wrote:

> +  // Align test coordinates to TWIP coordinate system and shift by a half
> +  // TWIP to avoid line junction situations which are hard to handle. 
> Oversample
> +  // everything by 100 to get some degree of accuracy (ie. this won't produce
> +  // any visible inaccuracy before the shape is scaled more an 2000x). The 
> +  // resulting coordinate is *very* close to the original one and still in 
> the
> +  // same coordinate system.

We'd need similar euristic for proper and performant withinSquareDistance
test for curves. The classic approach seems to be approximate and compute
distances, so knowing the actual rendering scale would help finding
an appropriate approximation. I guess this still needs a Gui handler
in the movie_root, or we'll miss the actual on-screen pixel scale...


reply via email to

[Prev in Thread] Current Thread [Next in Thread]