Re: Fwd: Dojinshi & Accessibility was: A new Semantic Web logo?

Hi, Guys-

Jonathan asked me to weigh in here.

~:'' ありがとうございました。 wrote (on 10/13/2007 2:30 AM):
> 
> Begin forwarded message:
> 
> Resent-From: semantic-web@w3.org
> From: Ivan Herman <ivan@w3.org>
> 
> Dear Jonathan,
> 
> let me give some backgroun explanation and propose a solution to see if
> that would work with you.
> 
> It is true that, in an ideal world, SVG should include texts as, well,
> texts, and not in geometry. And the specification of SVG allows that. It
> is also possible to add a description of a font (or the characters that
> are used from a font) into the SVG file itself to ensure a proper
> rendering on all environment, independently of the fonts being stored
> locally.

Indeed.


> The very practical problem, however, is that a number of SVG players
> (eg, Mozilla) do not understand (yet, hopefully!) those text statements
> with font descriptions. In order to have the logo displayed properly, we
> had to go  back to the original design, do a version that turned the
> characters into geometry, and export that into SVG.

A real-world problem.  FWIW, I think that in the next-next release of 
both Firefox and Safari, they will have partial support for SVG 
Fonts.... maybe a year, year-and-a-half way, hopefully.

In the meantime...


> However: what I would propose to do is to add RDF metadata into the SVG
> file (this is perfectly possible, too) that would include dc terms
> including a description containing the text of the logo (plus, eg,
> copyright info). Is this agreeable to you? This is something that could
> be done within a few days (alas, it should be done individually in a
> text editor:-)

Silly humans with their big brains.  Please find attached a moderately 
workable solution to the problem of a lack of SVG Font support.  Simply 
add in the Real Text, with a suitable-enough font to match the size and 
style of the logo, positioned more-or-less correctly, and set the 
opacity to zero.  Nothing shows to mar the nice rendering, but it has 
the text content, and it makes the faux-text seemingly selectable 
in-place and findable (and indexable).  This doesn't work well for 
dynamic content, but it's fine for static graphics.

I also added a <title> and a <desc> element with some suggested text.

This is not to diminish the inclusion of RDF, particularly with 
copyright info, etc.... but that's for a different purpose, I feel. 
Text intended to be human-readable (that is, rendered) in SVG should 
always be included using one of the text-content elements (<text>, 
<tspan>, <tref>, <textPath>, and <textArea>).

I'd like to see this done with all W3C logos, and I'm willing to put 
some cycles to it.

Regards-
-Doug Schepers
W3C Staff Contact, SVG, CDF, and WebAPI