Re: Formal definition of HTML5 (was Re: Version information)

At 01:54 PM 4/16/2007, Maciej Stachowiak wrote:

>Even in its current state, <canvas> probably defines the output more
>precisely than SVG.

This pixel-perfect discussion has left me a bit confused. Does 
Apple's patent claim on <canvas> cover any of the rendering methods, 
or is it that the patent covers exactly the <canvas> element as it 
appears in the WHATWG recommendation?

The reason I ask, is that I have noticed considerable quality 
differences in the way that VML for example is rendered, vs SVG with 
Adobe plugin in IE, vs SVG in Opera or Firefox. I naively assumed 
Adobe may have used patented rendering technologies in their SVG 
plugin and that was why it looked so good.

Why might that matter in this instance? Suppose that in order to 
actually build a suitable rendering engine for <canvas> as specified, 
that there might also be other IP entanglements, related to the 
rendering -- though not associated with the tag itself,  that might 
expose a manufacturer to risk? That is, might any extant patents 
limit a developer from being able to handle the pixel perfect stuff 
without exposure?

For example, in the extant implementations of <canvas> does Apple use 
any fancy QuickTime stuff, or are there gnu licenses or some such 
thing associated with Mozilla and Opera that would prevent another 
developer from having sufficient room to build a rendering engine 
from scratch without itself running afoul of someone's IP?

Speaking of IP entanglements, I would assume that ACD Systems 
(formerly Deneba) has stated that it has no problems with the use of 
"canvas" for this purpose, given their drawing software of that name? 
Seems like trademark dilution could rear its head if not.

Just curious,
David Dailey
http://srufaculty.sru.edu/david.dailey/svg/SVGAnimations.htm

Received on Monday, 16 April 2007 18:49:42 UTC