Re: Applying SVG properties to non-SVG content

On Sun, 13 Jul 2008 01:16:20 +0200, David Hyatt <> wrote:

> On Jul 11, 2008, at 1:41 PM, Robert O'Callahan wrote:
>> There's also a strategic issue within the standards community. It
>> seems the "path of least resistance" to adding a single feature to a
>> standard like CSS (or SVG) is to just go ahead and extend CSS (or
>> SVG) with a new spec for that feature, rather than creating a
>> dependence on another standard that has that feature.
> On the other hand if using the other standard would be clumsier than
> simply extending the original standard (as is the case with - for
> example - CSS-based gradients vs. being forced to link to an external
> SVG file just to do a gradient that is cleanly separated from your
> HTML presentationally), then I think that should be taken into
> consideration.

Sure, if you want to separate content into different parts then there are drawbacks like this. However, mixing the content (html+svg inline) also has other interesting benefits, and overcomes the concern that you raise, i.e. the need to reference (additional) external files.

> When integrating SVG and CSS we need to look for
> integration points that really unleash SVG's power into HTML (via
> CSS).  I think your proposal does this, so that's great.


> However I would not want to see this work used as an argument for
> preventing CSS from extending at all into territory that overlaps with
> SVG, since in many cases i think the CSS properties could integrate
> *back* into SVG (e.g., CSS transforms).

Indeed there are features in CSS that would be great to integrate into SVG in the future. 'text-overflow' is one example that has been requested in the past. And while I agree that some overlap may be inevitable, the overlaps in such cases should at minimum try to be compatible with each other.

I suppose css @font-face and <font-face> in svg is one example of having overlap, while still trying for the most part to be compatible.

> If the alternative would be extremely clumsy to specify in CSS, then I
> think integration with SVG is good. 

I'd be tempted to agree, but I think that it also needs to be extremely careful to not introduce incompatible definitions. If gradients are added to CSS I'd like to see some proposals for how they'd be meant to integrate and apply to SVG.


Erik Dahlstrom, Core Technology Developer, Opera Software
Co-Chair, W3C SVG Working Group
Personal blog:

Received on Monday, 14 July 2008 10:31:06 UTC