W3C home > Mailing lists > Public > www-svg@w3.org > November 2004

Re: Reconsider SVG 1.2

From: Robert O'Callahan <robert@ocallahan.org>
Date: Wed, 17 Nov 2004 19:58:04 -0500
Message-ID: <419BF39C.8060008@ocallahan.org>
To: Kurt Cagle <kurt@kurtcagle.net>
Cc: www-svg@w3.org

[The context was an off-list message in which Kurt advocated deprecating 
CSS and integrating all its features, including a full set of 
flowed-text features (floats, columns, etc), into SVG.]

Kurt Cagle wrote:

> I'm talking about interop, but not CSS interop. I don't think that 
> will happen, even if it does happen to be a good thing. CSS has a lot 
> of utility in HTML, and has some (albeit less) utility with rendering 
> straight XML. SVG does already incorporate many of the features of CSS 
> already, because both are presentation layers that support 
> container-oriented inheritance, though in many ways SVG is more full 
> featured than CSS because the semantics sit at a lower level of 
> abstraction.  When I reference interop, I'm talking more along the 
> lines of binding architectures (a la XForms), integration with XHTML, 
> X3D (or its equivalent), XML Events and the like.

You can't use XHTML without styling it somehow. Is XSL-FO your preferred 
mechanism for that?

>> I think you're completely wrong, and ignoring the benefits of 
>> leveraging the huge installed based of CSS renderers, authors and 
>> users. You're also advocating a major change in direction for the 
>> W3C. Nevertheless, please repost your message to the www-svg mailing 
>> list. It's important for everyone to know where everyone else stands.
>
> Frankly, I'm not advocating throwing away CSS. I'm just tired that all 
> of the CSS people seem to be so dead-set on trying to impose a pre-XML 
> standard with no pre-defined extensibility mechanism, no clean XML 
> parsing capability, the requirements of a secondary renderer and so 
> forth on SVG, simply for the sake of arguing about an installed base. 
> Within the SVG domain, there IS NO installed CSS base,

You're ignoring the use of SVG in mixed documents, especially HTML.

> and any CSS renderers that exist will have to be rewritten for SVG 
> anyway, regardless of what the final specifications end up looking like.

Peter and Jon keep claiming this is not so. I hope they're right.

> As to your comment about advocating a major change within the 
> direction of the W3C, I'm more puzzled than anything about that. What 
> radical change is that?

I had assumed that the W3C's new CDWG was tasked with making 
HTML+CSS+SVG work well, among other things, because this scenario was 
important to the W3C. You appear to believe that that is irrelevant 
because HTML and CSS are going away. Perhaps my assumption is wrong.

> Correct me if I'm wrong, but as a writer, developer, and evangelist of 
> this technology, I DO try to keep on top of what's going on, and have 
> since 1993. To me, CSS is a legacy application that needs to be 
> supported for the next several years because there is already a 
> reasonably large pre-existing base, but over time, it will fade in 
> favor of some XML based standard. I don't necessarily see that 
> standard being SVG -- there are aspects to presentation layers that 
> have nothing to do with the mandate of SVG -- but it will be something 
> XMLish (XBL, anyone?).

If the W3C is indeed deprecating CSS and HTML, then they need to fess 
up, terminate the CSS WG activity, hand those standards off to another 
standards body which is willing to evolve them, and let the marketplace 
decide.

If the W3C is not planning to deprecate CSS and HTML then we need to 
make it clear that SVG in HTML+CSS is an important use case, and if 
necessary, hold off on standardizing SVG 1.2 *for that scenario* until 
we have time (and implementation experience, and browser input) to get 
it right.

Rob

-- 
Robert O'Callahan <robert@ocallahan.org>
"Here is a trustworthy saying that deserves full acceptance: Christ
Jesus came into the world to save sinners-of whom I am the worst.
(1 Timothy 1:14-16)
Received on Thursday, 18 November 2004 00:49:28 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 5 February 2014 07:14:52 UTC