- From: David Hyatt <hyatt@apple.com>
- Date: Tue, 18 May 2010 19:36:08 -0500
- To: "Tab Atkins Jr." <jackalmage@gmail.com>
- Cc: Eli Morris-Heft <eli.morris.heft@gmail.com>, Alex Meiburg <timeroot.alex@gmail.com>, www-style@w3.org
On May 18, 2010, at 6:59 PM, Tab Atkins Jr. wrote: > On Tue, May 18, 2010 at 4:53 PM, Eli Morris-Heft > <eli.morris.heft@gmail.com> wrote: >> Would it be unacceptable to just say that, when using #rrggbbaa notation, >> the aa part maps 0x00-0xFF directly onto 0.0 - 1.0 (that is, take the aa >> value, convert it to decimal, and divide by 255), and if an author needs to >> specify a color with /exactly/ 0.5 opacity (or another float that is not >> represented exactly by one of the 256 values 0x00-0xFF) for some reason, he >> or she just has to use rgba() (or hsla()) notation? I think that most >> authors using (or rather, who would use) #rrggbbaa notation, and I am among >> this company, would be just fine with 0x80 representing 0.50196... > > That's what I'd prefer. Just dividing the value by 255 is the sane > way to handle it, and the fact that you can't specify exactly .5 is > pretty irrelevant. > > I'm just saying that our current webkit code is sorta crazy. If that's the craziest thing you've seen in the code so far, then you're in for some nasty shocks down the road. ;) dave (hyatt@apple.com)
Received on Wednesday, 19 May 2010 00:36:43 UTC