RE: Problems with guideline 4.1

Ah, ok. I can accept this approach, that if there is no public specification then the requirement to validate to spec is deemed inapplicable and therefore not a violation of guideline requirements.

I like Jason's suggestion that guidelines have preconditions that essentially state the scope of their applicability. Jason proposed a couple ways we might handle that, and I don't off hand have a preference among those or other ways that people might propose.

I think it would be useful to think of the preconditions slightly differently than what Jason stated though. Jason said that if a precondition is not applicable, the guideline should be treated as having been satisfied. I propose instead that if a precondition is not applicable, than the guideline is not applicable. This doesn't affect conformance to the guidelines materially, although possibly the conformance definition might need to be updated for this concept. But to me it would be a slightly more accurate way of stating your conformance. If you don't have non-text content, it would not be accurate to say you have met the requirement to provide text alternatives for non-text content. You haven't done that, because there was no non-text content for which to provide text alternatives. Instead, my proposal means you do not meet the requirement to provide text alternatives (since you did not actually do so) but that the requirement is not applicable due to the absence of non-text content, and therefore you can conform to the guidelines taken as a whole, even though you don't conform (nor do you fail to conform) to that particular guideline. For the hypothetical case of conformance claims on a guideline by guideline basis (e.g., an EARL conformance claim), this distinction would have some meaning, even though either way you still pass the conformance requirements.

Sorry if the wording of this makes the logic seem more convoluted than it is. I think it's actually pretty simple, and again does not make a material impact on the guidelines but to me makes an important logical impact.

Michael


> Correct. If there is a published specification, conform to 
> it; if there is
> no published specification then there is nothing to conform 
> to, and hence
> the guideline doesn't apply and should be treated as having 
> been satisfied
> by default, just as, where there is no non-text content in 
> the authored
> unit, guideline 1.1 should be taken as having been met.
> 
> Thus I am making two points:
> 
> 1. Guideline 4.1 only applies where there exists a published
> specification.
> 
> 2. As with other guidelines, if the preconditions for the 
> application of
> guideline 4.1 don't occur, the guideline should be treated as 
> satisfied.

Received on Thursday, 29 July 2004 10:23:25 UTC