------------------------------------------------------------------------ Date: Thu, 21 Apr 2005 14:20:34 -0400 From: "Weibel,Stu" Subject: Re: Definition Comments: cc: Liddy Nevile To: DC-USAGE@JISCMAIL.AC.UK ------------------------------------------------------------------------ I've had a private conversation with Tom about the definitions below, and with Tom's consent, I've offered to share my own observations. These remarks are public, open, and shareable with others, and of course brickbats are always welcome. I've corrected in the text the typos Tom told me about. 1. A description of the characteristics of a resource that may be relevant to the accessibility needs and preferences of a user. 2. A description of the characteristics of the resource which may enable or inhibit users' sensory perception or control of the resource or their intended engagement with the resource given their special needs in the context. In the interest of full disclosure, I'll say that Liddy shared with me an earlier version, and I made some recommendations that seem to have migrated (at least partially) into the first. The first one is not too bad in my estimate, though there are a couple words I'd still want to eliminate, but it is readable and, with the exception of the circularity of using the word 'accessibility' in the definition, understandable. The second definition I find overwrought. It is twice as long and half as clear. No one will know what it means. Approval of such a definition would, in my estimation, elevate confusion to new heights. I am given to understand that part of the problem of wording is to make the definition sufficiently abstract that things beyond information resources can be described with it (buildings, events, and such). This is probably a desireable goal... we've always tried to make DC as general as we can. But it substantially complicates the definition to do so. I don't have an immediate solution for this (alert the media) but I would like to give it a try, perhaps in a subsequent post. I do have a proposal about the underlying model for an accessibility term that might simplify the task. I know I have not been party to these discussions, and if my remarks are naïve or just wrong, ok, but I believe that you should not reject my model without clear examples of how it fails. My model: Accessibility issues take place in a tripartite context: resource ----- rendering device ----- agent domain A domain B I know that 'rendering device' is perhaps too info-asset-specific... I don't have a better term at the moment, but I'm confident that can be solved. My assertion is that the definition can be made far more succinct and clear if one avoids issues of 'user preference' and 'user capabilities' (domain B interactions). Accessibility metadata should not be constructed to address those issues, or cognitive issues, but rather to make it possible the matching of device capabilities and resource attributes (domain A interactions). Domain B is important, of course, but is accommodated by a marketplace that produces devices (screen readers, magnifying glasses, wheel chairs, canes, crutches, etc). The question for accessibility metadata is simply to make it possible for a system to identify appropriate resources that can be used in a given context. (building_has_wheelchair_ramps, Device_renders_HTML_to_voice, resource_encoded_according_to_ISO_12345, etc). So now for the brickbats....? stu ------------------------------------------------------------------------ Date: Thu, 21 Apr 2005 14:38:42 -0400 From: "Weibel,Stu" Subject: A variant proposal for an accessibility definition Comments: cc: Liddy Nevile To: DC-USAGE@JISCMAIL.AC.UK ------------------------------------------------------------------------ My model from my previous post: Accessibility issues take place in a tripartite context: resource ----- rendering device ----- agent domain A domain B Domain A being the technical matching domain, and domain B being the social and economic matching of devices between people and devices. In thinking about this, it occurs to me that the marketplace rarely, if ever, builds rendering devices without a clear specification to build to, nor are buildings built to accommodate arbitrary access issues, but rather are built to one or another legal specification. Why not structure accessibility metadata to reflect this? Its possible that some cases might be missed, but isn't it the case that what we are trying to promote is a declaration of conformance? This averts the problem of specifying a user's need/disability, and also the problem of finding a generalization for the 'rendering device'. The definition might then become: 1. A assertion of conformance of a resource to a public standard, specification, or legal requirement pertaining to accessibility. The acceptable values for such an element might include: - a formally maintained URI pointing to the standard, specification, or legal requirement (preferred). - the Title of the standard, specification, or legal requirement (less desireable, but probably still useful). ------------------------------------------------------------------------ Date: Thu, 21 Apr 2005 16:23:56 -0400 From: "Weibel,Stu" Subject: Re: A variant proposal for an accessibility definition Comments: cc: Liddy Nevile To: DC-USAGE@JISCMAIL.AC.UK ------------------------------------------------------------------------ Hi, Pete, Thanks for quickly responding to my modest proposal. I see the logic in your argument... It is homologous, at least to a degree, to the educational standards element. I personally have no considered opinion on whether this is in any important way different. The essence of my proposal is, I believe, agreeing on the model and promoting practice based on that. I am very curious as to whether the model is acceptable to the accessibility group. Indeed, my own understanding of the problem in this way originated with discussions I had with Liddy on the topic perhaps 2 years ago. So, I was surprised to find that it is not currently the way they are thinking. stu -----Original Message----- From: A mailing list for the Dublin Core Metadata Initiative's Usage Board [mailto:DC-USAGE@JISCMAIL.AC.UK] On Behalf Of Pete Johnston Sent: Thursday, April 21, 2005 4:12 PM To: DC-USAGE@JISCMAIL.AC.UK Subject: Re: A variant proposal for an accessibility definition Hi Stu, > The definition might then become: > > 1. A assertion of conformance of a resource to a public standard, > specification, or legal requirement pertaining to accessibility. > > The acceptable values for such an element might include: > - a formally maintained URI pointing to the standard, > specification, or legal requirement (preferred). > - the Title of the standard, specification, or legal requirement > (less desireable, but probably still useful). In the terms of the Abstract Model, that URI and that title would be a "value URI" and a "value string" respectively, and the "value" would be the standard/specification/legal requirement itself. However, taking the proposed definition and the comment about the values together, it is slightly ambiguous whether the proposal is that - the value is a standard - the value is an assertion of conformance to a standard I think there is a slight difference between those two things. If the value is a standard, and a statement made using the property dcterms:accessibility indicates an assertion of conformance to the standard (which I'm fairly sure is the intent of your proposal), then this seems pretty close to the existing property dcterms:conformsTo, defined as > A reference to an established standard to which the resource conforms. (which if it was revised to be more DCAM-compatible would drop the "A reference to" bit). The only difference is the qualification that in the dcterms:accessibility case it is a standard "pertaining to accessibility". If this is the case, is a new, separate property necessary? If a new property is necessary, then it seems to me it would be a subproperty of dcterms:conformsTo.... ------------------------------------------------------------------------ To: DC-USAGE-request@jiscmail.ac.uk From: Liddy Nevile Subject: Re: Accessibility definition Date: Fri, 29 Apr 2005 22:05:11 +1000 ------------------------------------------------------------------------ what Stu says next is not at all to the issue, I am afraid. The underlying model is not correct, sorry. > I do have a proposal about the underlying model for an accessibility > term that might simplify the task. I know I have not been party to > these discussions, and if my remarks are naïve or just wrong, ok, but > I believe that you should not reject my model without clear examples > of how it fails. > > My model: > > Accessibility issues take place in a tripartite context: > > resource ----- rendering device ----- agent > > domain A domain B > > I know that 'rendering device' is perhaps too info-asset-specific... I > don't have a better term at the moment, but I'm confident that can be > solved. This was an early idea about how to deal with accessibility and it does not work. > My assertion is that the definition can be made far more succinct and > clear if one avoids issues of 'user preference' and 'user > capabilities' (domain B interactions). Accessibility metadata should > not be constructed to address those issues, or cognitive issues, but > rather to make it possible the matching of device capabilities and > resource attributes (domain A interactions). Stu seems to be straying into device independence issues. Needs and preferences are vital in the accessibility context. People with disabilities often suffer from changes in their abilities due to such factors as tiredness etc and they also often have rights such as that they should have access to a dictionary that are then set aside in some special circs - eg when they are doing a spelling test. All these things have been worked out by people who understand the field and it is not a good idea to just guess what people with disabilities need. It is also not true that we don't need to think about cognitive disabilities etc - very smart people have car accidents and strokes and their cognitive capacity can be affected in a variety of ways - people with dyslexia etc also are considered to be people suffering disabilities. Such people need to be accommodated. > Domain B is important, of course, but is accommodated by a marketplace > that produces devices (screen readers, magnifying glasses, wheel > chairs, canes, crutches, etc). This is not the case.