W3C home > Mailing lists > Public > www-style@w3.org > August 2009

Re: [CSS3, Backgrounds and Borders Module] some questions about border-radius

From: Andrew Fedoniouk <news@terrainformatica.com>
Date: Sun, 23 Aug 2009 11:55:50 -0700
Message-ID: <4A9190B6.2030007@terrainformatica.com>
To: "Tab Atkins Jr." <jackalmage@gmail.com>
CC: Zack Weinberg <zweinberg@mozilla.com>, W3C Emailing list for WWW Style <www-style@w3.org>
Tab Atkins Jr. wrote:
> On Sun, Aug 23, 2009 at 1:14 PM, Andrew
> Fedoniouk<news@terrainformatica.com> wrote:
>> Here are three possible cases of the location of the transition area:
>> http://www.terrainformatica.com/w3/border-radius-transition-areas.png
>>
>> plus I've added for the reference example of Xara way of doing corner
>> transitions on wide borders. Algorithm is close to case (1) as far as I can
>> tell. At least it matches case #9 here
>> http://www.terrainformatica.com/w3/round-corners-sciter.png
>>
>> I suspect that current spec is trying to define case (1) also.
>>
>> Zack, your case is (3) and is different from what spec is saying
>> about tangents.
>>
>> We just need to choose one of cases define it.
> 
> I can assure you that I will *always* want case (3) rendering in that
> circumstance.  Both case 1 and case 2 end up with the horrendous
> effect that colors from my left border show up in what I consider my
> top border.  In your round-corners-sciter.png image, example 10 is
> horrifyingly bad and will *never* be the desired rendering of that
> code, believe me.
> 
> I'm not sure how to specify this without screwing up your other
> examples, which are all very attractive and pretty much exactly what I
> would want and expect (example 9 is weird, but understandable).
> 

All samples here:
http://www.terrainformatica.com/w3/round-corners-sciter.png
use the same algorithm, case(1).

If to use case(3) from classification defined in
http://www.terrainformatica.com/w3/border-radius-transition-areas.png
then it will break the spec as transition will happen outside areas
limited by quarter-ellipses of corners. That is the point.

-- 
Andrew Fedoniouk.

http://terrainformatica.com
Received on Sunday, 23 August 2009 18:56:24 GMT

This archive was generated by hypermail 2.3.1 : Tuesday, 26 March 2013 17:20:20 GMT