Re: [css3-background] border-radius color transitions using gradients recommended but undefined

Could you also live with #2? I kind of like the level of detail that's been put in of what is to be suggested, if a vendor is going to attempt it, in order to provide some minimal guidance. (as an aside, I would also add that on any border-radius in which the inner edge of the corner has an effective radius of zero, then the tip of the cone for the conic blend should align with that sharp corner.) I can also imagine some simple tests similar to what's been described here. But if we don't even mention the idea of a gradient blend at the corner, then we lose the ability to provide guidance (and a minimal level of consistency) to any implementor who wants to give it a shot.

On Feb 25, 2010, at 9:51 AM, Sylvain Galineau wrote:

> #3 is the simplest and clearest all around imo: the behavior is undefined. 
> 
> From: fantasai [mailto:fantasai.lists@inkedblade.net] 
> Sent: Wednesday, February 24, 2010 4:36 PM
> To: Sylvain Galineau
> Cc: www-style@w3.org
> Subject: Re: [css3-background] border-radius color transitions using gradients recommended but undefined
> 
> 
>   In the paragraph
>     # It is not defined what these transitions look like, but a gradient is recommended
>     # for color transitions that don't involve dotted or dashed borders.
>   replace 'recommended' with 'suggested'.
>     | It is not defined what these transitions look like, but a gradient is suggested
>     | for color transitions that don't involve dotted or dashed borders.
> 
> and Proposal 3 is
>   In the paragraph
>     # It is not defined what these transitions look like, but a gradient is recommended
>     # for color transitions that don't involve dotted or dashed borders.
>   replace the comma with a period and remove the rest of the sentence, i.e.
>     | It is not defined what these transitions look like.
> 
> Let me know if either 2 or 3 is satisfactory, and if so, we can take this to the WG for a formal decision. (If not, please explain what the remaining problems are.)
> 
> Thanks~
> 
> ~fantasai
> 
>