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

Hi, Ivan-

Ivan Herman wrote (on 10/13/2007 4:59 AM):
> 
> You are right with the silly humans:-): Adding a title and a desc is
> indeed another (and simpler) solution. Indeed, adding both is even better.

I still think that adding RDF would improve it.  The <title> and <desc> 
elements are relatively unstructured, and RDF could be used by a 
different processor to maximize the semantics.  In particular, specific 
copyright information conveys far more than the (R) ever could.


> However... Sorry to disappoint again with the Mozilla implementation,
> but it seems that your hack does not work either:-( I attach the screen
> dump of my Mozilla window. What does seem to work, however, is to set
> the text to visibility="hidden", instead of playing with the opacity
> value...

Doh!  That will teach me to monkey with a file (without re-testing) 
before I send it.  Indeed, it doesn't work with just 'opacity' (I'll 
report that bug); it does work with 'fill-opacity', though.  This is a 
better solution than 'visibility', because while Firefox doesn't let you 
select SVG text anyway, both Opera and ASV do... but not if you muck 
with the 'visibility' or 'display' properties; 'opacity' (while more 
computationally expensive) still puts it in the rendering tree, thus 
making it selectable.

Please find the corrected version attached.

I took the liberty of changing the file structure a bit, cleaning up 
some unused nesting levels of <g> elements, and reordering the 
glyph-shapes to match the visual order ('W3C' first, 'Semantic Web' 
second).  I grouped the individual <text> elements with their respective 
shapes (not necessary, but tidy), and added an 'id' for each shape with 
its character value (again, just a nicety, but might help if you decide 
to later use the shape when making an SVG Font).  I also put a 
<metadata> element in the (R) group, which I thought would be a clever 
place to put RDF relating to the copyright. :D

I'm sure that we all agree it behooves the W3C to demonstrate best 
practices whenever possible, and this is probably the most semantic logo 
possible, given current technology. :)  The only improvement I can think 
of would be to have a description of the box in the box group, either in 
<title> or <desc> elements (rather than just 'id' values), or in a 
metadata format.  But... I mean... there's only so far...

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