On Thu, 08 Feb 2007 12:04:43 +0530, Liddy Nevile <[log in to unmask]> wrote: > perhaps bear in mind, Charles, that we are not proposing a statement of > the type "x is accessible" or otherwise...but rather, "X has variable > font size" or "X is not controllable by keystrokes alone" Sure. But even in that context, how do I say "X has fixed font-size and I will do everything in my power to make sure X is an example of something whose font-size cannot be varied, because I am deliberately trying to make it bad", in a way that you can tell is different from "I don't know about the font-size in X"? (In EARL you could just say that it fails a particular requirement, but I am not sure how that fits in here). > Liddy > > On 08/02/2007, at 5:20 PM, Charles McCathieNevile wrote: > >> >> On Thu, 08 Feb 2007 01:54:25 +0530, Liddy Nevile >> <[log in to unmask]> >> wrote: >> >>> Emmanuelle >>> have you been able to read the metadata terms we are working on - >>> theyreflect the specs in WCAG but in a metadata way and give much >>> moreflexibility ... there is a fairly comprehensive set and they are >>> aboutto become an ISO standard at least for education.... >>> >>> The DC conformsTo etc simply did not give us enough useful >>> informationabout the resource..so we have added a few qualifiers and >>> one new term >> >> But the big question is whether you can state that something is not >> accessible (which >> is different from not stating that it is) - the use case is forexamples >> that >> are showing how not to do things. >> >> There is an issue if we use monotonic logic (which underpins the >> semantic web, a >> lot of the assumptions behind Dublin Core, and so on). Broadlyspeaking, >> a >> statement is expected to be true. If something is going to changestate, >> that is >> problematic, but there are ways around it. >> >> The reason for EARL's complexity is to provide a reasonable way of >> managing both >> trust, and changes in actual status fo the thing under discussion. If, >> instead >> of saying "foo is accessible" you say "fred says foo is accessible"then >> you don't >> make RDF and other monotonic systems crash if you add a statement "jo >> says foo >> is not accessible". (RDF doesn't actually ahve a "not" but you can use >> OWL to >> explain that passing "isAnInaccessibleExample" cannot happen toanything >> that >> meets WCAG-A, for example. You can then define"isAnInaccessibleExample" >> as test >> for EARL and make statements about it. >> >> As I understand it, you could then use that approach for DC metadata. >> >> So, as far as I know, there is no direct way of saying something is >> inaccessible, >> but it is easy enough to define a slightly indirect one. >> >> If anyone has a better approach I would love to hear it... >> >> cheers >> >> Chaals >> >> --Charles McCathieNevile Fundacion Sidar >> [log in to unmask] +61 409 134 136 http://www.sidar.org -- Charles McCathieNevile Fundacion Sidar [log in to unmask] +61 409 134 136 http://www.sidar.org