- From: David Dailey <ddailey@zoominternet.net>
- Date: Tue, 4 Feb 2014 13:35:01 -0500
- To: "'Paul LeBeau'" <paul.lebeau@gmail.com>
- Cc: "'Cameron McCormack'" <cam@mcc.id.au>, "'www-svg'" <www-svg@w3.org>
Received on Tuesday, 4 February 2014 18:35:32 UTC
I wrote: >> I'm not sure why we'd want to allow so poor an approximation as the bounding box of the control points (which could be very much looser than the curve itself). Paul wrote: >Could be, but in my experience it is not all that common that it has as major an effect as in your example. Plus there are also use cases where a fast and approximate bounding box is more desirable than a slow and accurate one. Perhaps one input to the algorithm might be a flag to indicate that you would be satisfied with an approximation. Hi Paul, If I understand the math here (it's questionable), it's not time consuming (linear, I think, in the number of control points in a path comprised of cubic and quadratic segments). See [1] for a nice little working demo where one drags control points and the BB adjusts. I also think that the case I presented is not at all unusual, nor extremal. See lots of different cases at [2]. Cheers David [1] http://floris.briolas.nl/floris/2009/10/bounding-box-of-cubic-bezier/ [2] http://www.math.ucla.edu/~baker/149.1.02w/handouts/bb_bezier.pdf
Received on Tuesday, 4 February 2014 18:35:32 UTC