RE: [css3-background] vastly different takes on "blur"

+1 vote for Brad Blur Ballot Bitmaps

• Brian

From: Brad Kemper [mailto:brad.kemper@gmail.com]
Sent: Tuesday, June 22, 2010 4:06 PM
To: fantasai
Cc: Brian Manthos; Simon Fraser; Tab Atkins Jr.; robert@ocallahan.org; www-style@w3.org
Subject: Re: [css3-background] vastly different takes on "blur"



On Jun 22, 2010, at 2:02 PM, fantasai <fantasai.lists@inkedblade.net<mailto:fantasai.lists@inkedblade.net>> wrote:
At least we're clear on what the two options are. If this is the
only complaint about the current definition, then I think we've
made a lot of progress on this feature!

Actually, at the risk of being even more hated, I think we should also consider the following, some of which is based on me thinking further about Brian's paint bucket experiment:

• should pixels that are >99% opacity or <1% opacity be considered as part of the transition area which is to fit within the defined space (if specified shadow color is 100%)? (my thought is "no", or else pick some smaaler threshold like >99.5% or <0.5%)


• if you don't like that, we should at least ask the question of whether or not full specified opacity or totally transparent can be considered to be part of the transition, or if they both (or either) live completely outside it. I don't think we are currently clear about that.


• should we say that all pixels of the blur that are <50% opacity (multiplied by specified color opacity) should exist outside the preblurred shape, and >50% within? I think implementations already do this, but I'm not sure how compatible it is with the other points, above.


As for whether the blur
value should match definition A or definition B, we've got two
options:
 1. Push the issue to the CSSWG and get a decision there
 2. Have Brad draw up some diagrams showing both interpretations
    for an offset shadow and a glow and post the question to
    css3.info<http://css3.info> / w3.org/bog/CSS<http://w3.org/bog/CSS> / twitter as a survey.

I'm willing.