W3C

- DRAFT -

ARIA APG TF

25 Jan 2016

See also: IRC log

Attendees

Present
IanPouncey, jongund, JamesNurthen, MattKing, AnnAbbott, CPandhi, TeresaBoers, JaEunJemmaku
Regrets
Jemma, Michiel, Bryan
Chair
SV_MEETING_CHAIR
Scribe
jamesn, matt_king

Contents


<jamesn> scribe: jamesn

Update on draft of landmark design pattern (John/Ann) https://rawgit.com/jongund/aria/master/practices/aria-practices.html#aria_landmark

MK: what state is this in - you sent an email

AA: not done yet but can take a look

JG: think it is greatly simplified
... added 2.1 and 2.2 sections

Sections 2.1 and 2.2 are "done" 2.3 is not

MK: perhaps folks should read it offline

https://rawgit.com/jongund/aria/master/practices/aria-practices.html#aria_landmark"

https://rawgit.com/jongund/aria/master/practices/aria-practices.html#aria_landmark

MK: one of things we talked about was rolling the content into example pages which demo the use of landmarks

AA: haven't started that yet

JG: interested in working on but nothing to show at this point

MK: since it is a WIP reviewing during the call may not be the most frutiful thing

JG: one question. The forms landmark. Put into APG that this is not a commonly used Landmark
... we were afraid that people might put a form landmark in

MK: I have never understood why we have it. Kind of like a cousin to application

AA: think it goes back to b4 ATs made big progress

MK: HPR revelaed form elements as they thought it was useful to the reader but no one does that any more

JN: potentially see the use in SVG or something

<mck> scribe: matt_king

<mck> IP: not sure we should discourage or promote use of form

<mck> aa: have to define it

<boerst> need to drop, due to conflict

<mck> jg: people see them; see form landmark, so use it because I have a form

<mck> iP: that's why we need to explain when to use it

<mck> jg: for search, we want people to use search landmark

<mck> jg: we need to show cases where it is useful

IP: same situation as navigation role. there are definate use cases where might want to naviagte

<mck> ip: I think it is similar to navigation; don't put on all linkks, but do so when it is significant

<mck> aa: consider login

<mck> jn: maybe a portal page where there is a small bit of the page for login, may be useful there

JN: some places forms may be ok.

Charu: are there any advantages?

<mck> cp: is there advantage of using it? we could list that?

<mck> aa: maybe the login example

<mck> jn: no disadvantage, we shouldn't steer people away

<mck> aa: concern people may use it for every form

<scribe> scribe: jamesn

MK: if the purpose of the form role was to have a mapping for the form element
... then I think we should make it a structure rather than a role
... dont think it is mapped by any browser
... another possibility to avoid that is that only forms with labels are mapped - like regions
... could use region from an AT perspective

JN: that goes for all landmarks really
... i disagree that form is less important

IP: making decisions about UI - i agree

MK: rare that there is a part of a web page which is not part of main which has a form
... even if it was a common practice the benefit is low. get the real info from the label not the type
... others get info with type

IP: can put form inside navigation
... no point in making structural
... dont make it structural

JN: is there any evidence this is a problem
... aria in HTML says <form> maps to role=form

JG: sometimes have empty forms in pages

so i think that is a bug

MK: any time there is an implicit role it puts it in the tree. If has role form will be regarded as a landmark

<jongund> https://www.w3.org/TR/html51/semantics.html#the-form-element

MK: exposing it is a decision of the screen reader.

JN: search is now allowed on form elements

MK: form is a landmark. unless we make a spec change need authoring guidance

JG: would a login form be an appropriate use of form

CP: looking at html spec. the form element had default saemantic as role=form. search and presentation are allowed

<jongund> https://www.w3.org/TR/html51/semantics.html#the-form-element

<cpandhi> https://www.w3.org/TR/html-aria/#index-aria-search

<cpandhi> https://www.w3.org/TR/html-aria/#docconformance

JG: primary content of authentication page is username and password

ARIA in HTML is a [HTML51] specification module. Any HTML features, conformance requirements, or terms that this specification module makes reference to, but does not explicitly define, are defined in the [HTML51] specification.

JN: is a rec track document
... very confusing documents

MK: wish we could fold "using aria in HTML" into the APG
... we should request a change.

JG: seemed to agree that was a good idea at the time

MK: from a perspective of minimizing role bloat - don't see a lot of value in the form role. Less to write abotu etc.
... wondering if should propose deprecating it in aria 1.1

JN: epub wants to get more specific on things

IP: for screen readers has no value unless used for navigation
... however, there could be future cases where it is useful

JN: what harm does it do if someone puts it in?

JG: perhaps describe as a specifialized region
... agree that the label on the form is more important than that there is a form there

MK: harm is overuse
... if used judiciously then it is not worth having region and form in my mind
... for 1 thing - in addition to dont use role form where there is an html form
... if a child of another region and the majority of the region is the form then don't use it

JN: should state that form elements do NOT have to be children of form role
... good uses bank sites have login regions etc.

example here - https://www.bankofamerica.com/

ARIA in HTML - https://www.w3.org/TR/html-aria/

<mck> https://www.w3.org/TR/html-aria/#docconformance

Update pattern work assignments and status https://github.com/w3c/aria/wiki/Aria-Authoring-Practices-Patterns-Status

Discuss example development work in progress

JG: will have menu bar update next week

<jemma> https://rawgit.com/jongund/oaa-examples/master/examples/menu-button/menu-button-1.html

JG: Is a menu button that links to other pages ok?

JN: a favourites menu would navigate to them

MK: if the links are persistant on the page - would not turn them into menu items.
... no one would be aware that could do other stuff. lose all the other semantics if you put it in a menu

just a set of links - dont put them in a menu bar

rather see them just in a navigation structure

JG: grid - was that for mega menus

MK: the element which opens the mega menu would move focus within it'

JG: would you constarin within megamenu until you close it

MK; yes or select something

JK: it opens the menu

JN: but focus doesnt move to the menu items
... you mean focus should move to the first item
... yes

MK: if tab to it and force jaws into forms mode then will track it - but if just press enter then it is as if the button did nothing as the focus didnt move
... good use of menu button Jon

JG: when you open a menu button focus should move to the 1st menu item?

MK: yes

AA: and esc should close

JK: it does

JG: should break the keyboard stuff up - for the menu and for the button
... also have someone working on combo box
... was close last week

MK: can add both to next weeks agenda
... probably a few more weeks on landmarks?

AA: probably

MK: combobox is 1 of the roles i'm working on for the spec
... there may be some things that dont quite work well yet. not going to disallow aria-owns. going to recommend aria-controls instead of owns.

Update pattern work assignments and status https://github.com/w3c/aria/wiki/Aria-Authoring-Practices-Patterns-Status

Review text of section 2.32 Tool Bar http://www.w3.org/TR/wai-aria-practices-1.1/#toolbar

http://w3c.github.io/aria/practices/aria-practices.html#toolbar

MK: in 1st para makes assumptions that there is a menu bar and it is a subset
... not sure that is true any more
... the assumption it manages focus is more of a practice
... not consistent

VO collapses toolbars and have to interact with them

Spec: "Authors may manage focus of descendants for all instances of this role, as described in Managing Focus."

"A collection of commonly used function buttons or controls represented in compact visual form."

Apple Definition (from OSX) - A toolbar (which is often combined with a title bar) gives users convenient access to the most frequently used commands and features in an app.

Oracle (ADF rich client) - Toolbars group iconic and textual commands which act on objects on a page or within a component (Tables, TreeTables and Trees). Toolbars contain at least one button along with optional separators, standard web widgets and overflow lists.

like here http://www.oracle.com/webfolder/technetwork/jet/uiComponents-toolbar-toolbarPattern.html

JK: what is the distinction between menu button and toolbar?

JN: can have a menu button in a toolbar

MK: will reduce description to 2 sentences from 3 paragraphs :)
... group info seems useless

JN: would agree with removing that

MK: RTL stuff? do we remove?

JN: I think we should have something general abotu all patterns which include left and right arrows

THE HOUR

OR 90 MINS

will pick up with the toolbars

Summary of Action Items

Summary of Resolutions

[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.144 (CVS log)
$Date: 2016/01/25 19:35:51 $

Scribe.perl diagnostic output

[Delete this section before finalizing the minutes.]
This is scribe.perl Revision: 1.144  of Date: 2015/11/17 08:39:34  
Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/

Guessing input format: RRSAgent_Text_Format (score 1.00)

Found Scribe: jamesn
Inferring ScribeNick: jamesn
Found Scribe: matt_king
Found Scribe: jamesn
Inferring ScribeNick: jamesn
Scribes: jamesn, matt_king
Present: IanPouncey jongund JamesNurthen MattKing AnnAbbott CPandhi TeresaBoers JaEunJemmaku
Regrets: Jemma Michiel Bryan

WARNING: No meeting chair found!
You should specify the meeting chair like this:
<dbooth> Chair: dbooth

Got date from IRC log name: 25 Jan 2016
Guessing minutes URL: http://www.w3.org/2016/01/25-aria-apg-minutes.html
People with action items: 

WARNING: Input appears to use implicit continuation lines.
You may need the "-implicitContinuations" option.


[End of scribe.perl diagnostic output]