https://www.w3.org/WAI/GL/task-forces/silver/wiki/Main_Page - all Silver prototype links
Shawn: How do we get more people
involved and how do we manage it?
... Our biggest challenge today is the Conformance model
Jeanne - we have 4 parts to COnformance: User Agent, Authoring Tool and AT guidance;
scribe: measurability
... point scoring system
... final score and levels
Shawn: Let's put the point
scoring system and the levels together
... we also have task based instead full-page
... We have to define what is accessibility and usability. It
may be a gray area.
... What do we need for the measurability?
Jeanne: We need more people with
experience in writing valid usability tests.
... we need more people with point score experience -- Audrey
who has experience with the French system. Anne with the
Norwegian experience, but they are working on ACT. We could ask
them to comment
... We should enroll Tim Boland from US and Simon Harper in the
UK. Simon is a researcher who has written papers on alternative
scoring systems.
... it's been working to have small groups, but maybe we need
to devote one meeting a week to the small groups instead of the
full group
Shawn: I would rather see the small group meeting when they need to.
jeanne: User testing
... we need to confirm the usability testing labs.
Shawn: We need to test the
feasibility and testability from the AGWG
... we may not get a lot of participants
... it's a small group to begin with, plus we have to get the
people who have the time.
... we have to go to the stakeholder map.
Jeanne: Let's do that today - go to the stakeholder map and figure out who needs to test the information architecture.
Shawn: FOr plain language, we want to have broad participation and tasks and comments. We need to get someone to help us set that up.
Jeanne: We should ask Dave Sloan to help design the plain language test. If he isn't available, he may be able to recommend someone.
Shawn: We could also ask Dave
about how to test information architecture.
... Regardless, we need plain language translations.
... we could write the basic information, and then turn it over
to the editors. We have a list of people who will help with
editing
... if we just draft things, and get halfway there, then the
editors can get it the rest of the way there.
Jeanne: We could ask AGWG for volunteers to write plain language following the template and the style guide.
Shawn: we could set it up in
Github,with an issue for each WCAG SC, then we can assign it to
people and track their progress. They can work in whatever tool
they want.
... Since Github is not accessible to everyone, we will only
use it for project management. The discussion will not be in
Github, it will be in the context of wherever the work is
happening.
... To clarify, we are talking about translation to plain
language.
... we also need to make a note of how to test translations of
the information architecture. That was a valid point raised
yesterday that there is a need for some identifier that is
cross language.
Jeanne: What is the use case for international translation and need of a unique number?
Reinaldo: I don't think we need a number. People care about the content, not the number.
[brief discussion of which WCAG SC to use as an example. 4.1.2 is too hard/big to start with.
<Lauriat> Draft doc: https://docs.google.com/document/d/18JyGF-AK8Qgq7DPyVlDYmxoj6814rORxuCf0l0oSb7U/edit
scribe: We are looking at 3.1.1 Language of Page as a starting example.
Wilco: Should there be individual Methods for each tests? I think a Method should have multiple rules, when appropriate. So for 3.1.1 has three atomic rules: HTML has lang attribute, HTMLLang and Xml:Lang Match; Validity of HTMLlang attribute; Specified Language matches reality (to be written).
Jeanne: Specified Language matches reality should be a manual test.
Wilco: We havne't written this one.
Shawn: The Guideline could be "The human language of the environment can be identified by assistive technology"
Jeanne: "Identify the human language for assistive technology."
Shawn: That sounds like you are setting Voiceover.
Wilco: Environment is defined for every technology
Shawn: Applicability of guidelines on technologies from ACT is a good way of identifying the gaps for the platforms (mostly user agent and assistive tech
[discussion of the ability to browser to expose the language to the user. Tools can detect it.
Shawn: If a browser and screenreader combination doesn't support a method, how do we identify that it?
In the example, we have 3 related metods: HTML tag has lang attribute
,,, HTTP Response Header specifies language
scribe: VR town has language
specified
... HTTP Response Header the browser can detect, but it doesn't
expose to assistive tech. This would be a method that is
desired, but itsn't supported by user agents.
Jeanne: Would ACT and Auto-WCAG be able to work on some of the tests to support Silver prototypes?
Wilco: I don't see why not as long as we stay inside the requirements of the funding.
Shawn: The biggest challenge was
taking it from HTML specific to generic of guideline
... we had questions about the plain language synonym of
"environment"
Amy: In most games, the language comes from the game engine
[the group looks at WCAG2ICT for examples of wording for technology neutral.]
Shawn: Can we have a combination of HTML tag and language on individual elements.
Wilco: Could we say that the content management system has to expose the language for the author to set?
Jeanne: That would be ATAG!
Add a related Method
Shawn: Give users a way to set a language attribute at the environment level.
Set a long description that describes clerly that it is for the authoring tool
<scribe> meeting: Silver TPAC meeting Day 2
This is scribe.perl Revision: 1.154 of Date: 2018/09/25 16:35:56 Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/ Guessing input format: Irssi_ISO8601_Log_Text_Format (score 1.00) Present: steve jeanne Shawn Audrey Cooper Wilco No ScribeNick specified. Guessing ScribeNick: jeanne Inferring Scribes: jeanne Found Date: 23 Oct 2018 People with action items: WARNING: IRC log location not specified! (You can ignore this warning if you do not want the generated minutes to contain a link to the original IRC log.)[End of scribe.perl diagnostic output]