minutes: UAWG telecon 23 May 2013

from: http://www.w3.org/2013/05/23-ua-minutes.html

User Agent Accessibility Guidelines Working Group Teleconference

23 May 2013

See also: IRC log
http://www.w3.org/2013/05/23-ua-irc


Attendees

Present
: 
kford,  Jim_Allan, Jan, Greg_Lowney, +1.609.734.aaaa, Jeanne, sharper,
Kim_PatchRegretsChairJimAllan, KellyFordScribeallanj, kford

Contents

Topics

Definition of browser - see "User Agent definition" thread
http://lists.w3.org/Archives/Public/w3c-wai-ua/2013AprJun/0056.html
Testing sub-group/writing tests

Summary of Action Items

________________________________

<trackbot> Date: 23 May 2013

<allanj> UAAG20 new version was officially published!!!

<allanj> scribe: allanj

open item 4

<kford> Scribe: kford

JA: Dates are 7/23-25

Group talking about dates briefly.

<jeanne> UAAG announcement tweet for those who want to RT ->
https://twitter.com/w3c_wai/status/337616376942428160

Definition of browser - see "User Agent definition" thread
http://lists.w3.org/Archives/Public/w3c-wai-ua/2013AprJun/0056.html

JA recaps thread.

<allanj> jan's comments
http://lists.w3.org/Archives/Public/w3c-wai-ua/2013AprJun/0057.html

JA: There were a couple of edits in there JAN. You said web view component
in mobile apps changed.
... There is a lot that is overkill in UAAG for our American Airlines app
example. In my new wording it would be covered.

Note, previous was from JR.

JA American airlines app, controls never change. In browser controls can
often change depending on content stream.

Jeanne: Stand alone web apps that get web content have no gu8idelines that
cover.

Group continues to discuss what stand alone apps should provide.

GL: Games do not want to provide a source view.

<Greg> (That is, most commercial webapps.)

Jeanne: If you can tell me the difference between a magazine app and the AA
app, let me know.

EH: Should there be a place for you could asssert that a certain SC would
be not applicable by virtual of the fact that this would fundamentally
alter the agent.

Jeanne: We do say if the platform doesn't support something the agent does
not have to do so.

EH: So you hare saying we already have th8is covered?

Jeanne: I think so but we'd need to test it.

EH: I sent an example of a testing application where you are very careful
about what you expose to the student.

<Greg> We should probably be updating and taking notes in
http://www.w3.org/WAI/UA/work/wiki/Use_Cases_for_UAAG_Applicability.

<allanj> jan: special case for constrained functionality (no code view for
testing application) would make a good note.

<allanj> eh: WCAG has an exemption from conformance for testing

<Jan> WCAG2 1.1.1 Non-text Content: All non-text content that is presented
to the user has a text alternative that serves the equivalent purpose,
except for the situations listed below. (Level A) - Test: If non-text
content is a test or exercise that would be invalid if presented in text,
then text alternatives at least provide descriptive identification of the
non-text content.

<allanj> jr: where to draw line between airline app and some other app with
a more detailed interface.

<allanj> current definition

<allanj> user agent

<allanj> A user agent is any software that retrieves, renders and
facilitates

<allanj> end user interaction with Web content. If the software only
performs

<allanj> these functions for time-based media, then the software is
typically

<allanj> referred to as a *media player*, otherwise, the more general

<allanj> designation *browser* is used. UAAG 2.0 identifies several user

<allanj> agent architectures:

<jeanne> ACTION: jeanne to work with Jan to work up language for the
glossary definition of user agent subtypes. [recorded in
http://www.w3.org/2013/05/23-ua-minutes.html#action01]

<trackbot> Created ACTION-831 - Work with Jan to work up language for the
glossary definition of user agent subtypes. [on Jeanne F Spellman - due
2013-05-30].

<allanj> stand-alone, non-web-based, browser: These user agents run on

<allanj> non-Web platforms (operating systems and cross-OS platforms, such
as

<allanj> Java) and perform content retrieval, rendering and end-user

<allanj> interaction facilitation themselves. (e.g. Firefox, IE, Chrome,
Opera).

<allanj> embedded user agent: These user agents "plug-in" to stand-alone
user

<allanj> agents in order to rendering and facilitate end-user interaction
for

<allanj> content types (e.g. multimedia), that the stand-alone user agent is

<allanj> not able to (e.g. Quicktime, Acrobat Reader, Shockwave). Embedded

<allanj> user agents establish direct connections with the platform (e.g.

<allanj> communication via platform accessibility services)

<allanj> web-based user agent: These user agents operate by (a) transforming

<allanj> the web content into a technology that the stand-alone (or
embedded)

<allanj> user agent can render and (b) injecting the user agent's own user

<allanj> interface functionality into the content to be rendered. (e.g.

<allanj> Gmail, Facebook, Skype)

<allanj> web view component, mobile app: These user agents are used to

<allanj> package a constrained set of web content into non-web-based

<allanj> applications, especially on mobile platforms. If the finished

<allanj> application retrieves, renders and facilitates end-user interaction

<allanj> with web content, then the application is a user agent. If the

<allanj> finished application only renders non web content, then the

<allanj> application is not a user agent for the purposes of UAAG 2.0

<allanj> conformance.

<allanj> scribe: allanj

better version

A user agent is any software that retrieves, renders and facilitates end
user interaction with Web content. If the software only performs these
functions for time-based media, then the software is typically referred to
as a *media player*, otherwise, the more general designation *browser* is
used. UAAG 2.0 identifies several user agent architectures:

stand-alone, non-web-based, browser: These user agents run on non-Web
platforms (operating systems and cross-OS platforms, such as Java) and
perform content retrieval, rendering and end-user interaction facilitation
themselves. (e.g. Firefox, IE, Chrome, Opera).

embedded user agent: These user agents "plug-in" to stand-alone user agents
in order to rendering and facilitate end-user interaction for content types
(e.g. multimedia), that the stand-alone user agent is not able to (e.g.
Quicktime, Acrobat Reader, Shockwave). Embedded user agents establish
direct connections with the platform (e.g. communication via platform
accessibility services)

web-based user agent: These user agents operate by (a) transforming the web
content into a technology that the stand-alone (or embedded) user agent can
render and (b) injecting the user agent's own user interface functionality
into the content to be rendered. (e.g. Gmail, Facebook, Skype)

web view component, mobile app: These user agents are used to package a
constrained set of web content into non-web-based applications, especially
on mobile platforms. If the finished application retrieves, renders and
facilitates end-user interaction with web content, then the application is
a user agent. If the finished application only renders non web content,
then the application is...

scribe: not a user agent for the purposes of UAAG 2.0 conformance.

<jeanne> Hotmail, Googledocs and Flexpaper for new examples of web-based
user agent

js: added mobile app handle to webview component

ja: web vs non-webcontent

jr: airline app is very constrained to its own data, all interface
components are html. perhaps only WCAG apply
... but there are constrained SC that do apply...like role, state, name,
etc.

<Jan> oops

js: perhaps the current definition sets too high a bar for web apps.

gl: reading over the definition...with a fresh look
... issue with lack of plugins and extension

js: plugins are embedded user agents

gl: still missing extension paragraph
... suggest spreadsheet of criteria that cause UA to meet the various
categories

jr: seems overkill
... an extension is nothing with out a UA to extend

eh: architectures implies there would be sections in the document. These
architectures only appear in the glossary.
... should be soffened, and not bolded (which implies that I can search the
doc for these terms, but nothing shows up)

jr: want to keep these. could see 'web view' could be a constrained
conformance claim
... new thinking.... android webview component could make a claim on the
component, not on a full browsers. but if you build a full browser based on
the component then have to meet all of UA
... webview would be a separate conformance claim

gl: confused by the wording. should apply to standalone and mobile apps.
... why only mobile.

jr: could have a desktop app.

gl: then just like any other program

jr: right, want to remove apps and focus on the component only.

gl: this really needs to be clear.

jr: the airline app would not need to make a UAAG conformance claim, only
the webview component would be a claimant

more examples/usecase discussion

jr: will ponder more, and work with JS to make things more

gl: will work with JS and JR

<Greg>
http://www.w3.org/WAI/UA/work/wiki/Use_Cases_for_UAAG_Applicability#Use_Cases_for_Conformance

Testing sub-group/writing tests

<kford> group first revisiting F2F schedule.

js: review test

sh: discussed his submissions, instructions for thought processes

js: these are good for constructing tests

ja: my example http://www.tsbvi.edu/uaag/1-3-1-test.html

jr: perhaps note the differences between the element (a-e) ensure they are
different, then look to see if they are changeable

eh: procedure to implement capability is different from testing a capability
... like SH precision available from pseudo-code. seemed a procedure for
implementing a feature rather than testing.

sh: yes, that's how I think about things...the code ends up being the
comments in real code.
... could write in rule based language (RAVEN, JESS - java expert system
shell)

<sharper> http://www.ibm.com/developerworks/library/os-eclipse-raven/

<sharper> 'Test GUI accessibility the Eclipse RAVEN way'

eh: sh code is a possible implementation. a rule for testing would fit many
implementations
... do we need to scope each SC, when do these get applied, ...

js: need to write assertions. for big items - how to test for wcag items.
included in part of the document.
... scoping is very important

jr: ATAG list test resources, a test document with all non-text content and
alternatives to test each sub SC.

js: remember, we will be doing the testing.

ja: should we list assertions/assumptions/preconditions

Summary of Action Items

[NEW] ACTION: jeanne to work with Jan to work up language for the glossary
definition of user agent subtypes. [recorded in
http://www.w3.org/2013/05/23-ua-minutes.html#action01]

[End of minutes]

--
Jim Allan, Accessibility Coordinator & Webmaster
Texas School for the Blind and Visually Impaired
1100 W. 45th St., Austin, Texas 78756
voice 512.206.9315    fax: 512.206.9264  http://www.tsbvi.edu/
"We shape our tools and thereafter our tools shape us." McLuhan, 1964