RE: Conformance use-cases

Hi Alastair,

I agree that the other A and B "Partial" versions just seem to complicate things. Re: IP and the Drupal case, the right to claim seems like it might be tied to the right to develop. The people with commit access to Drupal could surely designate someone to make a claim for them.

So here's an update of the idea:

1. Authoring System Conformance (A, AA, or AAA)
- can contain one or more* component authoring tools (*claimant is eligible to claim for the tool(s) or can point to the URI of a claim by an eligible claimant)
- the conformance claim can consider the tools together (if the claimant is eligible) or separately via their respective ""Partial" Component-Only Conformance" claims (see below). If considered seperately, the integration of the tools must not compromise the conformance of any of the constituents.
EXAMPLE: A developer claims this for a suite of integrated tools that meets all of ATAG Parts A and B to Level A.

2. "Partial" Component-Only Conformance (A, AA, AAA)
- for individual component authoring tool** (**claimant is eligible to claim for the tool)
- tool must meet Part A to at least Level A
- in Part B, SCs are recorded as Yes, No or Not Applicable (and the level is recorded for the multi-level SCs). 
- For any "No" answers, the tool must not prevent the SC from being met, in theory, by another authoring component as part of a larger authoring process. 
- It is allowed (but not required) that a URI for a conformance claim be provided for the other components that might be used to fulfill an authoring system conformance claim (see above).
EXAMPLE: A checking tool may claim "'Partial' Component-Only Conformance" for its checking and repair functions, while an editor without a checker claims "'Partial' Component-Only Conformance" for the rest of Part B. A service integrator then might claim "Authoring System Conformance (Level A)" by referencing the two "'Partial' Component-Only Conformance" claims and certifying that the integration preserves the conformance levels of the constituents.



Cheers,
Jan


> -----Original Message-----
> From: Alastair Campbell [mailto:acampbell@nomensa.com]
> Sent: November 7, 2011 5:08 PM
> To: w3c-wai-au@w3.org
> Subject: RE: Conformance use-cases
> 
> Hi Jan,
> 
> I think that's a good approach in general, although I think it is worth
> trying with just the full and component levels. Taking a fresh look,
> the partial component level should cover the A/B situations, shouldn't
> it?
> 
> More generally:
> For web content you test the web page/interface that the end-user
> encounters. It doesn't matter what back-end (or mashup) the site uses,
> it is what the user encounters that is tested.
> 
> For web authoring we need to test the interface(s) that the author uses
> to create content, and part of that test is on the web content that is
> produced.
> 
> What produces web content can be complex, and we should not try and
> predict exactly what that is. For me the key actor is the 'provider' of
> the interface that authors use.
> 
> I would like to make sure tool makers are encouraged to create a
> document that states:
> - Which ATAG SCs they meet.
> - Which SCs they do not try to implement (and of those, which can be
> met with a 3rd party tool).
> - Which SCs are dependant on how the tool is implemented (with notes).
> 
> BUT, it is then the provider of the final authoring interface who can
> create a full conformance claim. (Interface, workflow, solution, or
> whatever we call it.)
> 
> In the simple cases (1 & 6) the tool creator is the provider, so it is
> straightforward and they could make a full claim.
> 
> For the other cases, the tool makers would be making a partial
> (component?) claim. We should try to avoid a pass/fail mentality on
> partial claims.
> 
> Then, someone creating a CMS service that integrates other tools (case
> 2) can see what each provides, and make a full claim for their service.
> Someone creating service based solution (case 4) can pick a range of
> tools that allow them to create a solution that can make a full claim.
> 
> However, I don't think it is really possible to make a full claim until
> you have a complete (set of) tool(s) that an author would use. Somebody
> could easily mash together a set of accessible tools and loose all
> semblance of accessibility.
> 
> That does make it tricky for desktop products (case 3) aimed at end-
> users that do not make a full claim though, any ideas for that
> scenario?
> 
> Kind regards,
> 
> -Alastair
> 
> PS. I don't think that referring to IP is going to help. For example,
> who would make a claim for open source software like Drupal?

Received on Tuesday, 8 November 2011 03:26:59 UTC