- From: Charles McCathieNevile <charles@sidar.org>
- Date: Tue, 30 Sep 2003 16:50:10 -0700
- To: Jonathan Chetwynd <j.chetwynd@btinternet.com>
- Cc: wai-xtech@w3.org
Interesting questions... Some thoughts: You can't just add attributes to the metadata element in SVG. You could write a DTD for a derivative that does allow this, and write an XSLT or Xform that generates valid SVG using RDF inside the metadata element (and for that matter we could make it DTD-valid as well as schema-valid if we know in advance what kinds of metadata you are going to use). A few samples would indeed be helpful, and I wish I had some spare time. You could always hire me - I'm more expensive than a student intern though. Or search the Web for stuff... With your examples you have simple label (play) and a complex label (play video). You might use the desc element to provide a longer description of the functionality - it seems that the interaction model is only partially built into SVG, and there isn't the sort of "label", "help" stuff separated from information about the graphic components of the interface that you would get by integrating Xforms into SVG to build the interface... Alternatively, I don't think your particular example models very well what you are trying to do. It strongly suggests just using the desc element, but you might instead consider extending the use element (which is effectively an Xlink) with more stuff from Xlink, such as the role attribute... cheers Chaals On Tuesday, Sep 30, 2003, at 15:05 US/Pacific, Jonathan Chetwynd wrote: > > http://www.w3.org/TR/SVG11/metadata.html#Example wasn't that helpful > to me. > > Do people have examples of how to include (html type) meta > descriptions or keywords in SVG? is one allowed to? > RDF seems quite a large if useful, step to have to make. A few simple > RDF example graphics might also be helpful. > > What is the situation where a single graphic is being used in 2 or 3 > different contexts, and the meta tagging needs to reflect this. > > an example of the type of thing I have in mind is the single graphic > of a newscaster in 3 places: > > http://www.peepo.co.uk as a placeholder for our news > > http://www.peepo.com/alfi-x/reading.html for news to read > > http://www.peepo.com/alfi-x/tv.html for video news > > evidently it would be helpful if some metadata could be stored within > the <use> graphic, however probably additional material could be > stored on each page. > > To keep the coding short this example is intended to reflect the use > of a single graphic [triangle] to represent "play sound" or "play > video" dependent on context. > > <g > > <title>play</title> > <meta name="description" content="video" > <use xlink:href="../res/030831.svg#next" /> > </g> > > <g > > <title>play</title> > <meta name="description" content="sound" > <use xlink:href="../res/030831.svg#next" /> > </g> > > <g id="next"> > <meta name="description" content="play"> > <polygon fill="#cca" stroke="black" stroke-width="5" > points="55,10 55,140 115,75" /> > </g> > > evidently in this example a better title could suffice, however in the > real world it is necessary to provide extended search terms whilst > limiting titles, and for a given graphic it is likely that many search > terms will be common, but some will be context dependent. > > > My apologies if this is not as clear as it needs to be > > thanks > > Jonathan > http://www.peepo.co.uk > > -- Charles McCathieNevile Fundación Sidar charles@sidar.org http://www.sidar.org
Received on Tuesday, 30 September 2003 19:50:51 UTC