- From: <Anna.Zhuang@nokia.com>
- Date: Wed, 11 Feb 2009 08:56:05 +0100
- To: <public-pfwg-comments@w3.org>
Dear PF members, We in Education and Outreach group were actioned to review some of ARIA docs and I was adviced by Shawn Henry to send by bunch directly to PF list. That's what I am doing now, please have a look. Best regards, Anna Zhuang >-----Original Message----- >From: Zhuang Anna (Nokia-D-MSW/Tampere) >Sent: Friday, January 16, 2009 12:40 PM >To: 'ext Shawn Henry'; EOWG (E-mail) >Subject: RE: EOWG: Preparation for Friday 16 January 2009 >teleconference > .... > >ARIA Overview page does not contain WAI-ARIA User Agent >Implementation guideline. > >Generally on ARIA and WAI pages, I still don't see >relationship between WCAG and ARIA in any place. IMO ARIA is >no less important that WCAG. UAAG work hand in hand with WCAG >and now WAI-ARIA User Agent Implementation guideline has been >drafted. So this overall picture needs to be updated. > >WAI-ARIA 1.0 >============ >Abstract references ARIA Overview page. Should it instead give >the link directly to ARIA suite with or without mentioning >ARIA Overview? ARIA Overview is not meant for any specific >audience and reader of the overview may or may not read ARIA >spec. But once you are in ARIA spec, sure enough you are going >to read spec's introduction that provides information similar >to ARIA Overview page. > >Introduction has a sentence: "Assistive technology software >can transform the presentation into a format more suitable to >the user, and can allow the user to interact in different ways >than the author designed." Good to have some example here of >what suitable format may be and what would be different ways >of interaction. > >It is good to start Introduction with one paragraph of the >general problem statement (borrow something from ARIA >Overview). Othrwise Introduction jumps to what ARIA is >offering without explaining what are the problems that ARIA is >addressing. > >1.2 Use cases. One would expect a bullet list of use cases >here. The current material presentation hardly looks like use >cases, is not sistemised and not labelled in any way. > >3. I would change the title to ARIA Requirements. The current >title looks a bit odd from the table of content. Also each >chapter is anyway labelled as normative or informative. > >Can anything be done with the diagram in 4.4? If you print the >whole doc the diagram is unreadable. Even id diagram is scaled >and printed to A4 it is rather hard to concentrate on a very >small font. Place boxes closer and make them bigger? Stretch >diagram vertically so that it does not require so much >horizontal scrolling? > >Should definition of roles be moved to appendix and be >labelled as normative? > >No intro text between 5 and 5.1? > >States and roles are colour coded. I don't see the legend >explaining what is marked as yellow, grey or blue? Also is >colourcoding absolutely necessary? My experience is that it is >always good to have as black&white specification as possible. >Most people use BW printers to print large docs. > >WAI-ARIA 1.0 User Agent Implementation Guide >============================================ > >Not all links in the table of content work. Some jomp to the >top of the spec. > >The abstract is too short and this document it NOT YET part of >ARIA suite as described in ARIA Overview and noted by me >earlier in this mail. > >Introduction does not describe relations of this doc to ARIA >spec. generally the top of introduction lacks some general >text stating the problem this document is addressing. Also it >is important to state that the basics are explained in ARIA >and reader of this document is supposed to be familiar with ARIA spec. > >It would be good also to give a status of this document. It >does not seem to be a normative document. Perhaps can say that >this is supporting document to ARIA spec and the target >audience dor this doc is user agent developers. > >In the introduction it is good to have some reference to where >these APIs can be found ATK/AT-SPI on Linux >MSAA/IAccessible2 on Windows >At least MSAA has a reference link just before 4.3. need to >have it in the intro. > >Again colour coding does not have any legend. > >Last paragraph before 2.2. What is Gecko document? It is >neither linked nor explained. > >2.2 bullet 5. This is the first time AT occurs but the acronym >is not explained. > >2.3 title, replace AT with Assistive technology > >3. Mapping ARIA to APIs. Accessibility APIs? If so, it needs >to be stated in the title. This is important. > >Why managed stated deserve a separate paragraph and do not >come under the paragraph 3? > >Chapter 4 has no intro whatsoever. Oddly enough the title is >"managed states" but this chapter has zero occurrence of this >term. Is the title correct? > >WAI-ARIA Best Practices >======================= > >Unlike two other docs this one has a good abstract. This is >probably a credit to Lisa P. My only comment is, what is a >best practice? "This document specified best Practices..." >Best Practices is a title and there is a meaning to it. I >would replace the first sentence of the Abstract with the >first sentence from the Introduction (with slight >modification)--> "The WAI-ARIA Best Practices guide provides >readers with an understanding of how to use WAI-ARIA to create >an accessible Rich Internet Application." > >Also the word "provide" is used too frequently. > >Chapter 2 mentions "existing ARIA-enabled widget library". I >wonder if such libraries exist publicly? I would remove the >word "existing" alltogether. The message will not suffer any loss. > >Again colour coding has no legend. > >Unify the titles, either providing/setting/allowing across the >doc or provide/set/allow. Now it is a mix. > >In chapter 9 what does it mean "Placeholder for Local Index"? > >Typo in the first sentence of chapter 10. > >In chapter 10 I would remove two sentences that have >recommendations to reuse existing libraries. Firstly it is a >matter of ones choice to reuse someone's work or not. >Secondly, being true to the spirit of public standards, some >existing implementations should not be promoted by the >standards. I would write chapter 10 in the following way: > >Writing rich internet applications is much more difficult than >writing in HTML. It requires extra effort to ensure your >application runs in multiple browsers and supports WAI-ARIA. >Authors of rich internet application widget libraries will >have to go through: > >* extensive testing with assistive technologies >* cross browser testing >* testing to ensure that the widgets respond to desktop settings >* testing to ensure that the widgets match a common keyboard >style guide > >There are publicly available UI component libraries which have >already implemented WAI-ARIA, e.g. Dojo Toolkit Dijit library. >(Any other? Who can guarantee quality assurance of Dojo Tookit >to comfortably reference it in W3C standard???) Reusing such >libraries is one way to start developing accessible rich >internet applications. > >That's all I've got after scrolling 3 docs. > >Regards, >Anna > >
Received on Wednesday, 11 February 2009 07:57:09 UTC