- From: Leonard Rosenthol <lrosenth@adobe.com>
- Date: Sun, 5 Aug 2012 15:30:50 -0700
- To: Benjamin Hawkes-Lewis <bhawkeslewis@googlemail.com>
- CC: Sam Ruby <rubys@intertwingly.net>, "public-html@w3.org" <public-html@w3.org>
On 8/5/12 1:54 PM, "Benjamin Hawkes-Lewis" <bhawkeslewis@googlemail.com> wrote: >Do you mean that there are two profiles of conformance for PDF/UA, a >full set of rules to be validated by humans (perhaps with machine >assistance) in which the value of Alt is checked (however >subjectively) and a second profile that formally subsets those rules >for machine-only checking? Technically, there is only a single profile - as described in the standard itself. That profile, however, includes items that can be machine checked and items that can only be human checked. So that to TRULY comply with the standard, it MUST undergo human validation. It is impossible (by design, today) to completely machine validate a PDF/UA file. >an Alt attribute that does >not conform to common sense (e.g. a photo of an apple described as a >photo of a house) would nevertheless be fully conforming to the >standard? Of course not. Such a description would fail the human test, which is required for the VALUE of Alt. >\HTML and WCAG alike currently have some conformance requirements that >are not machine-checkable and other requirements that are, but they >don't formally designate which is which. Neither does PDF/UA. Leonard
Received on Sunday, 5 August 2012 22:31:23 UTC