- From: Chris Lilley <chris@w3.org>
- Date: Thu, 30 Mar 2006 21:02:08 +0200
- To: T Rowley <tor@cs.brown.edu>
- Cc: Jon Ferraiolo <jonf@adobe.com>, www-svg@w3.org
On Thursday, March 30, 2006, 8:11:29 PM, T wrote: TR> On 1/24/06 11:18 PM, Jon Ferraiolo wrote: >> Here is how that section of the spec now reads: >> ----------------- >> 11.3.3 System Paint Servers >> >> The following list of System Paint Servers must be supported. If a paint >> specification specifies one of the System Paint Servers, then the user >> agent must either paint using a system-provided paint server or paint >> with a substitute paint server, such as a color or gradient. System >> paint servers often depend on the operating system, user choices, and >> the implementation. Substitute paint servers should attempt to match the >> appearance of corresponding user interface elements on the platform, >> including user color choices. In environments which do not provide >> adequate system paint server APIs, a conformant user agent may use >> substitute paint servers which do not necessarily match the >> environment's system paint servers. >> ----------------- TR> This new language still allows system paint servers to provide TR> gradients, Yes TR> but without rules about what sort of space they live in (or TR> spreadMethod) it won't be possible to write interoperable content that TR> uses them. The actual gradient is not accessible. As the spec makes clear, unlike other paint servers where the computed value is an IRI, for system paint servers the computed value is the system paint name (like 'ButtonFace', for example). The stop colors, positions, opacities, and other values that define the gradient are not accessible to the author or to scripts. They are black box paint servers. By their very nature they are platform-specific in appearance. Your point about spreadMethod is a good one though- I would expect such servers to use the object bounding box, not a coordinate system; so spreadMethod would not have any effect. It might be worth mentioning that object bounding box is expected. But then, maybe some platform uses repeating gradients for that stripey semi-transparent plastic look. It seems better to let the implementation choose a suitable server internally. -- Chris Lilley mailto:chris@w3.org Chair, W3C SVG Working Group W3C Graphics Activity Lead Co-Chair, W3C Hypertext CG
Received on Thursday, 30 March 2006 19:02:11 UTC