Re: [CSSWG] Minutes and Resolutions 2009-02-04: box-shadow and border-image

My objection to building shadows into the border-image itself (and I  
say this knowing full well that we do this on OS X right now, e.g.,  
iChat), is that the shadows then become part of the the object's  
border box.

We have shadows built in to border images in iChat, and we have all  
this negative margin trickery to simulate the fact that the shadows  
should be pushed out beyond the border.  It's gross.

If the issue is having full control of what a shadow looks like, and  
being able to specify some kind of image for the shadow, then why  
don't we just extend box-shadow to support images?

dave
(hyatt@apple.com)

On Feb 12, 2009, at 2:05 PM, Brad Kemper wrote:

>
>
> On Thu, Feb 12, 2009 at 11:20 AM, David Hyatt <hyatt@apple.com> wrote:
> I don't agree with the idea that border-image should suppress box- 
> shadow, but I also don't agree with the idea of drawing a box-shadow  
> that follows border lines/curves that you don't even plan to  
> display.  If box-shadow used the border-image as a mask when  
> drawing, then that seems like the best solution.
>
> Then I think that we probably just differ on what is best and what  
> is second best. To me, "drawing a box-shadow that follows border  
> lines/curves that you don't even plan to display", as happens in  
> current implementations, would be the worst solution. So would you  
> agree that having the border-image suppress the box-shadow would be  
> better than the current implementations, and thus your second choice  
> if (for whatever reason) you couldn't have your first choice of  
> using the border-image as a mask for dynamically generated shadows?
>
> For me, choosing between those top two choices, I'm more in favor of  
> the authorial control of setting the shadow in the image editing  
> program that authors would use for the images anyway. I find the  
> arguments against it (that the shadow having the same image  
> resolution as the border image is a bad thing, or that dynamically  
> changing the color or size or angle of the shadow without loading  
> another image, or that having the clickable area no worse than it is  
> for current authors using background images to create their shadows,  
> or that it is hard to create convincing shadows in SVG using the  
> tools that authors would likely use) to be unconvincing, and of very  
> little merit.
>
> So here is how I would rate the choices:
>
> 1/2 star: drawing a box-shadow that follows border lines/curves that  
> you don't even plan to display
>
> 4 stars: box-shadow used the border-image as a mask (if the issue of  
> transparency can be dealt with in a fairly reasonable way)
>
> 5 stars: let the artist creating the raster border images also  
> create the raster shadows, and suppress the border-box-following box- 
> shadow in the same way that the padding-box-following borders are  
> suppressed.
>
>