ISSUE-2286: Resolving relative values in @kernelUnitLength (number-optional-number)

resolve_kernelunitlength

Resolving relative values in @kernelUnitLength (number-optional-number)

State:

RAISED

Product:

Module: Filters

Raised by:

Erik Dahlström

Opened on:

2009-06-29

Description:

When resolving relative values based on @primitiveUnits for @kernelUnitLength it's unclear what result is expected.

Example:
<filter primitiveUnits="objectBoundingBox">
<feConvolveMatrix kernelUnitLength="8" .../>
</filter>

a) @kernelUnitLength is expanded to "8 8", then relative values are calculated
b) relative values calculated for the x value, then duplicated to y

The kernelUnit will be slightly weird if the b) option is picked because it's then not depending on the height of the target element boundingbox at all. Thus I'm leaning towards clarifying that a) is the expected result.

[1] http://www.w3.org/TR/SVG11/filters.html#feConvolveMatrixElementKernelUnitLengthAttribute

Related Actions Items:

No related actions

Related emails:

1. Minutes, July 7 2009 SVG WG telcon (from anthony.grasso@cisra.canon.com.au on 2009-07-08)
2. Fw: Agenda, July 7 2009 SVG WG telcon (from cam@mcc.id.au on 2009-07-07)
3. Agenda, July 7 2009 SVG WG telcon (from cam@mcc.id.au on 2009-07-07)
4. Agenda, July 1 2009 SVG WG telcon (from ed@opera.com on 2009-06-30)
5. Agenda, July 1 2009 SVG WG telcon (from ed@opera.com on 2009-06-30)
6. ISSUE-2286 (resolve_kernelunitlength): Resolving relative values in @kernelUnitLength (number-optional-number) [SVG Filters 1.2] (from sysbot+tracker@w3.org on 2009-06-29)

Related notes:

No additional notes.

Display change log