It would be more direct to use coloralphafrominternal, which returns [24-bit RGB, alpha fraction]. If you trace inttocolorobj, it is scaling the alpha by 100 and you are unscaling it by 100, you could cut out the middle man and save 2 function calls... I'm just saying'.
On 2010-12-07, at 20:23, Henry Minsky wrote: > Change hqm-20101207-QEn by [email protected] on 2010-12-07 23:09:59 EST > in /Users/hqm/openlaszlo/trunk3 > for http://svn.openlaszlo.org/openlaszlo/trunk > > Summary: fixes for swf10 shadow rendering > > New Features: > > Bugs Fixed: LPP-9494, LPP-9565 > > Technical Reviewer: max > QA Reviewer: ffeng > Doc Reviewer: (pending) > > Documentation: > > Release Notes: > > Overview: > > > Details: > > + shadowcolor alpha value needs to be converted to value between 0 and 1 > > + for LzSprite, create a dedicated sprite to form a filled bounding > box for the shadow filter shape. > > There may be an optimization to use the bgsprite as a shadow outline > if it exists and is hidden, but I didn't want to try to figure out the > logic for that right now, and it's not clear it would work if alpha != > 1. > > Tests: > > For swf10, verify that the focus shadow glow effect applies to > bounding box of the whole component (button), not to the internal text > or icon. > > + demos/house.lzx > + demos/ptw.lzx > + see test case attached to bug report, lpp-9440.lzx > > Files: > M WEB-INF/lps/lfc/kernel/swf9/LzTextSprite.as > M WEB-INF/lps/lfc/kernel/swf9/LzSprite.as > > > Changeset: http://svn.openlaszlo.org/openlaszlo/patches/hqm-20101207-QEn.tar
