- From: R.J.Koppes <rikkert@rikkertkoppes.com>
- Date: Wed, 24 Aug 2005 16:53:00 +0200
- To: "Ian Hickson" <ian@hixie.ch>, "Emrah BASKAYA" <emrahbaskaya@hesido.com>
- Cc: <adelfino@gmail.com>, <www-style@w3.org>
I agree with the anti-aliasing thing. Maybe, we could rephrase it by saying that in case of rounding and a 50% decision occurs, the latter rule wins. This is where differences between browsers occur. Rikkert Koppes http://www.rikkertkoppes.com ----- Original Message ----- From: "Emrah BASKAYA" <emrahbaskaya@hesido.com> To: "Ian Hickson" <ian@hixie.ch>; "R.J.Koppes" <rikkert@rikkertkoppes.com> Cc: <adelfino@gmail.com>; <www-style@w3.org> Sent: Wednesday, August 24, 2005 4:30 PM Subject: Re: [CSS3 Backgrounds and Borders Module] : 1 pixel corner color > On Wed, 24 Aug 2005 17:07:54 +0300, Ian Hickson <ian@hixie.ch> wrote: > > > > > On Wed, 24 Aug 2005, R.J.Koppes wrote: > >> > >> everytime a device pixel has to "choose" between two colors, the latter > >> rule is applied. First check which border occupies the most space of the > >> device pixel, if both are equal, apply the latter rule. > > > > This seems to preclude high-quality anti-aliasing. We definitely don't > > want to prevent UAs from making the Web page look even better. > > > > What's the benefit in limiting what approximations a UA has to make when > > there are device limitations? > > > > I agree, if we were to set a pixel based rule, this would not bode well > with anti-aliasing. Anti-aliasing UA's will surely do sub-pixel sampling > and it would not need a rule. Non-anti aliasing UA's would simply round > things anyway, like they do know. > > > -- > Emrah BASKAYA > www.hesido.com >
Received on Wednesday, 24 August 2005 14:52:41 UTC