Fwd: review of ATAG 2.0 WG Draft 20-Jan-04

>Subject: review of ATAG 2.0 WG Draft 20-Jan-04
>To:	Jutta Treviranus <jutta.treviranus@utoronto.ca>
>From:	Barry Feigenbaum <feigenba@us.ibm.com>
>Date:	Mon, 2 Feb 2004 15:29:41 -0600
>X-PMX-Version: 4.1.1.86173
>Status:  
>
>
>
>
>
>Not sure how to submit these via normal channels.  Sorry they are too late
>for today's discussion.
>
>[I need to be added (or find out how to add myself) to the normal
>distribution list (so far Kip has had forwarded me most notes).]
>
>
>My comments on this document:
>
>General:
>
>This document seems easier to read than the 1.0 version.  Congratulations!.
>
>I believe the use of "must" and "all" in many criteria is too strong
>(especially with respect to my comments on 1.5 below).    Less all
>inclusive options should (or should that be "must") be available.
>
>I may come up with more comments later.
>
>Specific:
>
>1.3 [ATAG20-CHECKLIST] link seems to point to a nonexistent doc
>
>1.5 Conformance
>
>I think just the three A levels are too restrictive.  Its sort of
>all-or-nothing.  Tool authors cannot get partial credit.  For example, what
>if a tool is AAA for guideline 1 but not even A for Guideline 2:- What is
>its rating? - nothing!  What if its AAA for all but one of the priority 1
>sub-criteria of a guidelines, which it misses - What is its rating?
>
>I believe there should be a way to get credit on a guideline-by-guideline
>basis.   Within a guideline what if only some sub criteria [say 1.x] (but
>not all) at a given priority are met. What credit can be claimed (today its
>seems like 0 credit).
>
>I propose a set of achievement levels (something like):
>A - as today, all items of the criteria are met
>B - meets the spirit of the criteria (all possible met, good cause ** why
>others not met), and/or is WCAG friendly
>C - minimal achievement (some criteria meet, improvement is possible)
>
>So some tool could be rated 1-A, 2-B, 3-AA, 4-A (in form GL#-Level, ...) .
>This could make creating a certification logo a little problematic. Perhaps
>we could offer a simple tool to create the image.
>
>The more we can allow tool makers to take credit for what their tools can
>do (and let them advertize it) , the more likely tools will meet more
>criteria.
>
>** often for practical economic and/or schedule constraint reasons
>
>On "bundled" - what does this mean exactly? (a glossary term perhaps)  And
>"must be distributed together" seems quite onerous. Consider Eclipse.  Many
>plug-in are bundled (see IBM's WSSD, WSAD, Rational tools, etc) into
>products.  But others are sold/distributed stand-alone.  Yet each is
>expected to play well in the sandbox (look similar, behave similarly and
>interact with each other).
>
>2 Guidelines comments
>
>GL 1.1 I think we need a more specific SW Accessibility Guidelines (SWAG)
>reference (if we are trying to be normative).  The techniques doc provides
>some detail here but its not (I believe) normative.
>
>GL 1.3 criteria 1 "All" too strong (suggest "At least 1")
>
>GL 1.4 criterial.  "move/select/cut/copy/paste/etc" how?  Maybe defined in
>SWAG, but undefined here.
>
>GL 1.4 criteria 1 "all ... views" too strong (suggest "At least one")
>
>GL 2.1 "is valid for"- what does "valid" mean in this context (conforms to
>DTD, schema, or what?), how can the tool author verify it?
>
>GL 2.1 criteria 1 "applicable" - latest level only or any approved level?
>
>GL 2.4 criteria 1 "must" - what if the tool (often a third party one)
>doesn't do this?
>
>GL 2.6 criteria 1 May be exceedingly difficult to achieve for some required
>types (such as applets, ActiveX objects, etc) especially if they are black
>boxes
>
>GL 3.1 criteria 1 This seems to require the tool to understand and VALIDATE
>all content (vs treating it as a black box)  Hard to do.
>
>GL 3.2/3.3 criteria 1 "must" too strong again.  Many useful tools will not
>have checkers and/or fixers.
>
>GL 3.7 criteria 1 "All" - what if there is no help system (ie only hardcopy
>docs)
>
>GL 3.8 criteria 1 "must meet ... WCAG" - what if the sample screens aren't
>web content (ie WCAG does not apply).
>
>GL 4. 2 criteria 4 "all accessibility-related" - too strong
>
>
>I have some typos to point out:
>
>GL 1.1 "editor [to] display preferences"
>
>GL 3.3 "factors [should | must] be considered"
>
>Barry A. Feigenbaum, Ph. D.
>Worldwide Accessibility Center - IBM Research
>www.ibm.com/able,
>w3.austin.ibm.com/~snsinfo
>voice 512-838-4763/tl678-4763
>fax 512-838-9367/0330
>cell 512-799-9182
>feigenba@us.ibm.com
>Mailstop 904/5F-021
>11400 Burnet Rd., Austin TX 78758
>
>Sun Certified Java Programmer, Developer & Architect
>IBM Certified VisualAge C++ & Smalltalk Associate Developer; XML Developer;
>OOAD w/UML
>Brainbench in C/S, WWW, e-Commerce & OO Concepts, HTML 3.2, Java 2, JSP,
>EJB, XML, Smalltalk, Assembler
>
>This message sent with 100% recycled electrons


--

Received on Monday, 2 February 2004 16:38:14 UTC