- From: Jeanne Spellman <jspellman@spellmanconsulting.com>
- Date: Tue, 10 Mar 2020 22:40:40 -0400
- To: Silver Task Force <public-silver@w3.org>
- Message-ID: <7033285f-18ab-1124-0047-9b7abd45da6e@spellmanconsulting.com>
We had a snafu with the IRC logging (the old UTC midnight problem), and
some of the minutes were not captured in an attractive way. We do have
the logs, however, and I have pasted the relevant log into this email,
for those who want to read the missing minutes.
== Summary ==
ACT Tests: How do we structure our ask of ACT? What from the existing
ACT work can we bring over to 3.0? What can we expand upon? Can we
leverage the Applicability section of ACT for use in Silver Task
Completion testing?
* ACT Rules Format: https://www.w3.org/TR/act-rules-format/
* ACT Rules: https://act-rules.github.io/rules/
We want to schedule several joint calls with ACT, now that we have
content that we can work with. Prior to this, we would meet, but we
couldn't get far without content. We may also want to include people
familiar with the WCAG-EM tool. We reviewed Rules structure and the
WCAG-EM tool (which covers scope, sampling, and reporting -- not
testing). The rules are WCAG agnostic and cannot be filtered by SC, but
can be page-searched for SC number. They are technology specific and
appear to be mostly HTML and ARIA.
Task Completion Testing: How to create a framework for Task Completion
Testing? We started with definitions and discussion of different terms
to use. We want overall test results better to reflect people with
disabilities experience based on what they're trying to do. We discussed
scope and clarified that tasks cannot be scope, only logical subsections
can be a scope for purposes of conformance. We agreed that there can be
tasks in the scope, but the task cannot be the scope. Some cautioned
that we have to be careful not to allow companies to put in a phone
number to call instead of web accessibility. Should we base it on the
standard development life cycle? Some said that they let the site owner
define the task, and the tester only cares about the functional
outcome. We need to define a list of what the user would be doing to
apply toward conformance, so it can be repeatable. UXPA Usability Body
of Knowledge uses high-level tasks, broken down to subtasks - referenced
on usability.gov
We want to include the path of approach (for example, a login process
before performing the task) as part of the whole task completion. There
may be alternate paths that may be more accessible to some than others.
We also need to talk discovery of any alternate paths. Repeatability is
important for those with legal responsibilities. Scalability is
important to industry. We want to support both users and content creators.
Challenges document: We are being asked if we have questions we want
asked as part of publishing the Challenges
<https://raw.githack.com/w3c/wcag/conformance-challenges-redesign2/conformance-challenges/index.html>document.
Challenges is being published by AGWG, not Silver, but the feedback that
Challenges gets may be useful to Silver. What questions would we want to
ask? There is a list of existing questions. There was off-topic
discussion about the Challenges document. There was a suggestion that
it be reformed as a Roadmap document. Two questions were suggested:
* Given the challenges we have identified, what is the priority that
the reviewers think we should address them in?
* Are there specific success criteria in the appendix that should be
significantly revised and in what priority?
== Minutes ==
(missing from the start of Part 2 to the middle of the Task Completion
discussion. See the Text of Minutes section following this section for
the IRC log <https://www.w3.org/2020/03/10-silver-irc> of the missing
discussion. )
https://www.w3.org/2020/03/11-silver-minutes.html
=== Text of Minutes ===
agendum 4. "ACT Tests" taken up [from jeanne]
23:03:49 [sajkaj]
present+
23:04:24 [sajkaj]
sl: How to structure our ask of ACT
23:04:56 [sajkaj]
sl: What from the existing ACT work can we bring over to 3.0?
23:05:14 [sajkaj]
sl: What can we expand upon?
23:05:24 [JF]
Q+
23:05:40 [sajkaj]
sl: ACT has an applicability concept, as do we in Silver.
23:05:40 [Jennie_Delisi]
Jennie_Delisi has joined #silver
23:06:04 [Jennie_Delisi]
present+
23:06:20 [sajkaj]
sl: Understand that it's highly specific on a per rule basis and
described for how it applies
23:06:33 [david-macdonald]
david-macdonald has joined #silver
23:06:52 [sajkaj]
sl: We will certainly need to facilitate people tunneling into
guidance that pertains
23:07:05 [jcraig]
present+
23:07:06 [sajkaj]
sl: It should also facilitate us exposing tech gaps
23:07:11 [CharlesHall]
CharlesHall has joined #silver
23:07:18 [sajkaj]
sl: Restating page for apps, vr, etc
23:07:20 [CharlesHall]
present+
23:08:02 [kirkwood]
present+
23:08:04 [sajkaj]
sl: A gap might be if platform didn't support lang definition;
23:08:28 [sajkaj]
sl: Expect we have much to teach each other
23:08:41 [jeanne]
ACT Rules Format: https://www.w3.org/TR/act-rules-format/
23:08:41 [Lauriat]
ack JF
23:09:04 [sajkaj]
jf: ACT rules format might template some of our other testing
23:09:21 [sajkaj]
jf: Believe it would even apply to a cognitive walkthrough
23:09:42 [sajkaj]
jf: Because we're asking a series of yes/no questions in a walkthrough
23:09:45 [sajkaj]
sl: Agrees
23:09:57 [JennC]
JennC has joined #silver
23:10:18 [sajkaj]
sl: Some rules are automatable and others aren't, but structure
supports repeatability
23:10:30 [sajkaj]
jf: Critical
23:10:54 [sajkaj]
sl: They've done lots of work since we last talked, we need to know
23:10:57 [jeanne]
q+ to say that some tests will be yes/no, but not to require all
tests to be yes/no.
23:11:05 [sajkaj]
sl: Next is to learn what we can reuse and map that out
23:11:12 [Lauriat]
ack jeanne
23:11:12 [Zakim]
jeanne, you wanted to say that some tests will be yes/no, but not to
require all tests to be yes/no.
23:11:36 [sajkaj]
js: We should have yes/no wherever appropriate, but not all our
tests will be binary
23:12:03 [CharlesHall]
act rule design: https://act-rules.github.io/pages/design/rule-design/
23:12:10 [sajkaj]
sl: Don't believe ACT inherently requires binary; believe it guides
through making judgement calls
23:12:29 [sajkaj]
sl: Ex: alt for whether it provides equivalent experience
23:12:48 [sajkaj]
sl: How well, how close to equivalent might be scaled
23:12:51 [david-macdonald_]
david-macdonald_ has joined #silver
23:12:52 [sajkaj]
jf +1
23:13:09 [sajkaj]
jf: Even complex questions can be broken down to simple binary
23:13:22 [sajkaj]
jf: So, it's how we structure and string together
23:13:33 [jeanne]
Not all comple questions can be broken down to simple binary and get
valid results
23:13:46 [jeanne]
s/Not all comple /Not all complex
23:13:52 [JF]
CAn you provide an example Jeanne?
23:14:22 [sajkaj]
sl: Believe we can reuse both very granular but also more in depth
exploration toward test results
23:14:52 [jeanne]
@JF, Simplified navigation comes immediately to mind
23:15:00 [CharlesHall]
+1 to using the rule design
23:15:47 [sajkaj]
lucy: Are we stopped because we have no ACT people on the call?
23:15:57 [sajkaj]
sl; If you're in ACT, please speak up
23:16:18 [sajkaj]
lucy: Notes Mary Jo has been regular participant in Silver recently
23:16:53 [sajkaj]
lucy: Walking through their questions without their assistance has
been a profitable exercise for me
23:17:11 [sajkaj]
sl: We have an example, but it's quite old
23:17:27 [sajkaj]
js: Let's schedule a joint call with them.
23:18:09 [sajkaj]
js: In the past we were stopped in our joint attempts because we had
no content, we do now
23:18:14 [Chuck]
q+ point of order that need not be scribed.
23:18:33 [sajkaj]
sl: Probably schedule a series of calls
23:18:59 [CharlesHall]
q+
23:19:00 [david-macdonald_]
q+
23:19:09 [Lauriat]
ack CharlesHall
23:19:38 [sajkaj]
ch: Would be worth scheduling to have someone familiar with EM
reporting tool
23:19:48 [sajkaj]
js: Know whom to ask
23:19:55 [Lauriat]
ack david-macdonald_
23:20:28 [sajkaj]
dm: Working with aCT on Canadian Gov study ...
23:20:40 [sajkaj]
dm: I can talk some about that ...
23:20:53 [sajkaj]
dm: There are about 60 rules now applying to 2.x SC
23:21:02 [sajkaj]
dm: about 10 test cases, very granular
23:21:50 [sajkaj]
dm: Applicability is whether the test case is applicable to the rule
23:22:08 [Rachael]
present+
23:22:57 [sajkaj]
dm: Ex test would be about filename
23:23:27 [sajkaj]
dm: then what rule is applicable, expected outcome, whether it passes
23:23:44 [sajkaj]
dm: They are trying to build automatable tests
23:23:55 [JennC]
q+
23:24:07 [sajkaj]
dm: Mostly Wilco adding tests
23:24:11 [Lauriat]
ack JennC
23:24:24 [sajkaj]
zakim, who's here?
23:24:24 [Zakim]
Present: jeanne, sajkaj, ChrisLoiselle, Laura, Jennie, kirkwood,
Lauriat, Lucy, alastairc, Makoto, Chuck, JF, stevelee, KimD, AndyS,
PeterKorn, mattg, Rachael, Detlev, Fazio,
23:24:28 [Zakim]
... Jennie_Delisi, jcraig, CharlesHall
23:24:28 [Zakim]
On IRC I see david-macdonald_, JennC, CharlesHall, Jennie_Delisi,
KimD, Chuck, JF, mattg, Fazio, AndyS, RRSAgent, maryjom__, Zakim,
jon_avila, jeanne, Lauriat, kirkwood, jcraig,
23:24:28 [Zakim]
... sajkaj, MarcJohlic, MichaelC, Rachael, achraf, alastairc, yatil,
AWK, trackbot
23:25:06 [JennC]
https://act-rules.github.io/rules/
23:25:10 [sajkaj]
jc: Have a pointer to this
23:25:27 [david-macdonald_]
https://act-rules.github.io/testcases.json
23:27:01 [sajkaj]
sl: re alt-text == filename -- what's the logic?
23:27:15 [sajkaj]
dm: If you start on JC's page ...
23:27:50 [david-macdonald_]
https://act-rules.github.io/rules/
23:28:13 [sajkaj]
dm: Contains the rules -- about 60
23:28:20 [sajkaj]
dm: Click on context
23:28:22 [Lauriat]
https://act-rules.github.io/rules/ff89c9
23:28:47 [Chuck]
q+
23:28:52 [sajkaj]
dm: Will get dail incl linkage
23:29:05 [sajkaj]
lucy: Are they filterable by SC?
23:29:07 [sajkaj]
dm: No
23:29:13 [sajkaj]
dm: They're wcag agnostic
23:29:28 [sajkaj]
dm: They borrow from multiple standards
23:30:13 [sajkaj]
js: Can you talk us through Image has Accessible-Name?
23:30:29 [jeanne]
https://act-rules.github.io/rules/23a2a8
23:30:57 [Chuck]
q?
23:31:30 [jeanne]
ack chuck
23:31:42 [sajkaj]
ch: These seem ml specific?
23:31:45 [sajkaj]
dm: For now
23:31:49 [JF]
+1 yes
23:32:12 [sajkaj]
dm: Effort is to get all the automated tools gathered on one page
23:32:16 [JennC]
+1 yes for the purpose: harmonisation
23:33:09 [sajkaj]
dm: Responding to trying to get consistent rendering despite
browser's tendency to render differently
23:33:37 [sajkaj]
sj: Quips it's also very today, not just the 1990's
23:34:06 [Lauriat]
Applicability: The rule applies to HTML img elements or any HTML
element with the semantic role of img that is included in the
accessibility tree.
23:34:38 [Chuck]
q+ lucy
23:35:06 [Lauriat]
ack lucy
23:35:33 [sajkaj]
lucy: Do I understand we're asking whether their struct is
applicable to our needs? Not necessarily the actual rules?
23:35:47 [sajkaj]
lucy: Believe we want their structuring
23:35:50 [sajkaj]
sl: Agreed
23:36:11 [sajkaj]
sl: Expect we will be able to reuse their rule content
23:36:33 [CharlesHall]
+1 to use of rule design
23:36:46 [sajkaj]
sl: Hope is that we'll be able to reuse
23:37:05 [PeterKorn]
PeterKorn has joined #silver
23:37:08 [PeterKorn]
present+
23:37:28 [jeanne]
WCAG-EM
23:37:29 [CharlesHall]
WCAG-EM Report Tool
23:37:35 [david-macdonald_]
Website Accessibility Evaluation Report Generator
23:37:52 [david-macdonald_]
https://www.w3.org/WAI/eval/report-tool/#!/#%2Fopen
23:38:33 [sajkaj]
dm: Walks through the form ... ...
23:38:48 [sajkaj]
dm: Notes it's highly 2.x based
23:39:39 [sajkaj]
dm: It helps you gather up structured samples and also random samples
23:39:55 [sajkaj]
dm: At the end a report
23:40:01 [jeanne]
This is a tool for the spec that we based representative sampling on.
23:40:20 [sajkaj]
sl: Walks one through generating a report, but not through the
actual tests/
23:40:22 [sajkaj]
dm: yes
23:40:47 [sajkaj]
js: It's a tool for the spec that we based Silver conformance on
23:41:12 [sajkaj]
js: That's why it should sound familiar! We've been talking about
the sampling part
23:42:38 [sajkaj]
sl: Shifting to how to create a framework for defining a task and
scoring task completion
23:43:01 [sajkaj]
sl: Perhaps we might discuss how we might do it, what it means to
get this task defined
23:43:19 [sajkaj]
sl: Maybe there are better terms, but this is our best to date
23:43:39 [sajkaj]
sl: We want overall test results better to reflect pwd experience
based on what they're trying to do
23:44:07 [sajkaj]
sl: A tiny image in the corner that does nothing and has no alt text
is not really an impediment
23:44:23 [sajkaj]
sl: On the other hand the image that is a buttn and brings up the
menu is a big deal without alt
23:44:25 [CharlesHall]
useful resource on how to identify top tasks (analysis)
https://alistapart.com/article/what-really-matters-focusing-on-top-tasks/
23:45:01 [sajkaj]
sl: Using we've been saying "tasks" is to cover tech other than just web
23:45:21 [sajkaj]
sl: Pizza shop: find how to call; hours; menu; etc
23:45:40 [sajkaj]
sl: For Google Docs there are tasks about formatting, sharing, etc
23:45:53 [PeterKorn]
q+
23:45:57 [Chuck]
q+
23:46:11 [sajkaj]
sl: How do we guide people to defining their tasks. How do we help
NY Times do this?
23:46:19 [Lauriat]
ack PeterKorn
23:47:08 [Fazio]
Likert scale
23:47:10 [PeterKorn]
q-
23:47:15 [sajkaj]
pk: Jumps back to unlabeled icon out of the task flow -- want to
make sure we agenda defining what pieces matter, and which don't;
should help us realign with clear lang, COGA, etc
23:47:49 [sajkaj]
sl: Agree, but let's hold off for now until after the definition.
But, yes, it also applies to defining scope for conformance and may
depend on a particular path through the page, for instance
23:48:07 [Lauriat]
ack Chuck
23:48:10 [sajkaj]
sl: there may be parts of the page that interfere--things in the
environment that arn't directly related
23:48:18 [Lauriat]
q+ Lucy
23:48:29 [sajkaj]
ch: Should we base on standard development life cycle?
23:48:32 [Fazio]
That's the first question we ask UX research participants: "How
would you expect to perform this task"
23:48:50 [sajkaj]
ch: Would it pertain to sw dev?
23:48:58 [Fazio]
Then we walk them through our intended process flow
23:49:40 [Lauriat]
ack Lucy
23:49:44 [sajkaj]
s/ch/ca/
23:49:48 [sajkaj]
ch: maybe
23:50:06 [sajkaj]
lucy: we let the owner define how to build the task; we only care
about functional outcome
23:50:33 [sajkaj]
lucy: believe our job is to define success
23:50:51 [sajkaj]
sl: But we need to define what we need in order to explain what
we're asking people to do
23:51:12 [sajkaj]
sl: Person running the task is perhaps not the person who defined
the task
23:51:20 [sajkaj]
sl: One may need to explain to a colleague
23:51:53 [sajkaj]
lucy: OK, just don't want us to get bogged down by what should be
accomplished by one visit to http://123.456
23:52:22 [sajkaj]
sl: Need to define a list of what the user would be doing, in order
to define conformance
23:52:36 [sajkaj]
lucy: prefer actions and/or activities to the term task
23:53:10 [sajkaj]
ch: Think I understand the concern incl with the vocabulary; but
there's history here
23:53:30 [sajkaj]
ch: Author defines scope -- a better term is fine -- but it's a
common term in the field
23:53:38 [jeanne]
q+ to clarify scope
23:54:29 [Fazio]
q+
23:54:57 [sajkaj]
sl: We need a way to explain what blocks a user doing some thing on
a page -- we need to be able to explain this thing here is a big
problem, but this over here is annoying and time consuming, but not
a showstopper
23:55:15 [sajkaj]
sl: Really don't like to go more granular because that takes us back
to element based
23:55:37 [sajkaj]
sl: e.g. there may be three ways to accomplish a particular task,
which may require different action paths
23:57:01 [Jennie_Delisi]
UXPA Usability Body of Knowledge uses high-level tasks, broken down
to subtasks - referenced on usability.gov
23:57:06 [Lauriat]
ack jeanne
23:57:06 [Zakim]
jeanne, you wanted to clarify scope
23:57:06 [sajkaj]
ch: If I recall you can't start at an arbitrary point ...
23:57:28 [sajkaj]
js: We had considerable discussion about scoping
23:57:50 [sajkaj]
js: It has to be a logical subset, but can have various tasks, but
it's not the scope itself
23:57:51 [Lauriat]
ack Fazio
23:58:32 [sajkaj]
df: Do a lot of tasks with user testing, when I work with orgs we
work through the steps involved in achieving a particular outcome --
a task
23:59:04 [CharlesHall]
+1 human task(s) are about getting to the outcome
23:59:06 [sajkaj]
df: Having the author define this puts the onus on the author to
identify their functional outcomes
[1]W3C
[1] http://www.w3.org/
- DRAFT -
Silver Virtual F2F Tuesday Part 2
10 Mar 2020
Attendees
Present
KimD, Jennie_Delisi, PeterKorn, CharlesHall, AndySomers,
Lauriat, Fazio_, Rachael, sajkaj, JF, jeanne, kirkwood
Regrets
Chair
Shawn, jeanne
Scribe
kirkwood, Rachael
Contents
* [2]Topics
1. [3]Setting up a framework for task completion testing
2. [4]What feedback would Silver like to ask for from
publishing the Challenges document?
* [5]Summary of Action Items
* [6]Summary of Resolutions
__________________________________________________________
<jeanne> scribe: kirkwood
Jeanne: befor call starteddavidtalking about a resource
DavidF: usability testing versus accessibility testing posted
to slide share
<Fazio>
[7]https://www.slideshare.net/DavidFazio1/designing-for-users-w
ith-disabilities-presentation-135931732
[7] https://www.slideshare.net/DavidFazio1/designing-for-users-with-disabilities-presentation-135931732
DF: slide 14 is where the word document is
Jeanne: i can’t interact with it
SL: downloadable version?
DF: I can email
Jeanne: put in silver folder
JF: put somewhere we can look at it.
DF: silver folder good idea
<jeanne>
[8]https://drive.google.com/open?id=1j7gY-jwEn4vp2hPCJqEArufqmP
XJgGj6
[8] https://drive.google.com/open?id=1j7gY-jwEn4vp2hPCJqEArufqmPXJgGj6
Jeanne: I’ll get you one
LG: what is the main poihnt
SL: How we can set people up to create these tasks for scope
... setting up framework for tsk completion setting
tsk/task
SL: criticality would be part of this conversation
... from getting from point A to Point B rules around heading
where diret path may only bring you only through one part of
heading struxture
... just part of a complet heading structure
... if overall is a bunch of nonsense may cause issue
... animations you cant stop or flashing at bad rate even
though separte still part of envirnoment preventing you from
moving forward
JF: underscoredc current concern areound scoping. we can’t do
component testing because it is sum of components
... concerned about scoping creating rosier then reality
SL: what do need to take into account
JS: the scope can’t be the tasks
... heres the NYTimes ap for you we are testing
... the scope has to be a logical subsection of the product
... taks are part of scipe
JF: individual components will pass example given
JS: scope is a logical subsection of product
Janina: need to take whole county rich and poor
JF: starts at H3 all H# but skippin H2 all fails
SL: not talking components in isolation
LG: scares me, alternative path is the path of travel. we know
by dominoes this doesn’t pass
... can be gamed.
Jeanne: not saying thaty
<Chuck> early regrets. I must go.
SL: for an assessment need to do we need to define what users
are trying to do and what outcome is
... how do we create a structure so we can guide, assess, and
express results of test
... so we can ensure we aren’t testing compnents in isolation
... rather than real experience
LG: need to have an equal experience not alternative experience
SL: not sure how this applies here
LG: scape to read article not experience
SL: thats exactly what we want form first entry point every
path navigate to arrive at it
LG: each path should be equaivlent
SL: something we can talk about diffent paths such as making
bold in google docs test ability on all paths to make
acce3ssibile such as voice command which might not be
accessibile to everyone
<CharlesHall> i think of path to (approach) versus path through
(actions). as similar to multiple ways.
LG: not all methods to take paths may exist
SL: not all equivlent
... keyboard shortcut example to select bold versus voice
amount of steps
LG: check how applied to see how succeed. not significnatly
differnt
SL: i think we are actually agreeing
LG: also need to talk discovery
SL: back to top level discussion
... mulitple paths and steps to completion
... need to be careful about not chaing to technology, device,
applicaiton websit
LG: using recognized techniques not a recognizable techniques
Jennie: one challenge, when one has completed a set of tests
and vlidadates or retests need to use sme measurement of scope
especially with legal. repeatablity is very important
... would be concerned that 3 people testing would have
differnet results
LG: use path of travel in scope
Jennie: complaint comes into my office saying this issue is not
adddress. where would path exist for those to retest
LG: think need to expand scope
SL: ro add to point of repeatability
ro/to
SL: waling through decision tree are many ways to accomplish
task
DF: WCAG and Silver are diffenernt conformance odels are we
moving away from that?
SL: diffent conformandcde model than 2.x yes that is corrent.
but repeatability has subjectifvity and get away from attmepts
of black and white. through equivalency
... discussed keyboard equivalency
DF: never going to have black and white answers, judgement
calls frustrating or complet barrier
SL: we have that doay
<CharlesHall> shawn made my point
SL: we have these types of situations not just pass fali
... this is passing but could do better
... trying to get language for that
JF: I’d be careful about stance, can’t scale for industry
... if we make it hardr to do evaluation
... won’t be adopted
SL: repeatability of tests themselves
<JF> we need 3 things: measurability, testability,
repeatability
SL: repeatbility of test results
... hope we can narrow down
LG: gesture to complicated
JF: if we can't repeat test can't sue
<Fazio> +1 JF
LG: the result could have happended. keystroke 3exqample. test
was repeatable the results wernt
JF: if you cn’t prove an assertqation after fact don’t have a
leg to stand on
... support users and content creators
<KimD> +1
SL: if you produce certian test results, we should have
evidence on how to get results, target deviation for differnet
results
... structure to provide evidence for results
<KimD> +1 because the different results could be because 1
tester used JAWS and the other NVDA
LG: originally was tested for one group not another
<JF> Exactly Jennie
Jennie: from a content creator, and emplyee with diwsability
side. issue gets brought to court has to hear argument being
sid from both siedes and make a judgement. to see if testing
done is sufficient i’m concerned language isn’t going to
support this. need to make it tight for all parties involved.
... for example alt text yes/no
... where we get grey is the quality of alternative test
test/text
<CharlesHall> +to thresholds not being grey area
SL: exactly what we are trying to do
JF: if we don’t emerge with a standard that the indudstry cn
use they wont pick it up
... need stick not just a bag o carrots
... needs to work for everyone
SL: we absolutely do
... we need to make sure usability of standard is a major
consideration
... for strucrture scope and tasks we need to break down to
create structrue
... meeting with ACT group is imortant
... to define how to go through testing as well
DM: my understanding one stpe away form ACT theres EM framework
is not same crowd
... seven step model of defining scope
... don’t think ACT has been setting up an audit
SL: not what i was saying
... can reuse dstructrue framework to apply at higher level to
wlak through process of soing higher level t3est and
specifiying ceope
ceope/scope
SL: can use at a higher level
DM: can look at process
Jeanne: written proposals are welcome
SL: doesn’t need to be comprehensive
... shoot us an email
Setting up a framework for task completion testing
What feedback would Silver like to ask for from publishing the
Challenges document?
Janina: peter on?
PK: yes
<sajkaj> Latest of Challenges is here:
<sajkaj>
[9]https://raw.githack.com/w3c/wcag/conformance-challenges-rede
sign2/conformance-challenges/index.html
[9] https://raw.githack.com/w3c/wcag/conformance-challenges-redesign2/conformance-challenges/index.html
PK: thank you. basically we are hoping to soon get review
approval to publish first wokring darft of challenges document
... what specific questions would like to see accompany first
working draft for ongoing wok
<sajkaj> Here come the questions ...
<sajkaj> 1. Are there additional challenges not described in
our document?
<sajkaj> + Are there additional Success Criteria from WCAG 2.x
that should be added to the list of challenging SC in Appendix
A?
PK: questions are have we found all of challenges what is
missing, misstate wrong inaccureate
<sajkaj> + Are there additional considerations that should be
noted or corected among the SC described in Appendix A?
<sajkaj> + Are there similarly additional Success Criteria, or
additions or corrections to those enumerated in Appendix B that
should be noted, or otherwise described when applying WCAG 2.x
SC to ICT?
<sajkaj> + Are the enumerated challenges in Appendix A and/or
Appendix B suggesting certain patterns that should be called
out and described?
<sajkaj> + Are there aspects of our document that are incorrect
or insufficiently defined?
<sajkaj> + Have we overlooked some aspect in our analysis that
should be addressed?
<sajkaj> 2. What approaches to you believe are most promissing
to explore in constructing a more comprehensive conformance
model for WCAG 3.0?
<sajkaj> 3. What do you believe are the most promising
approaches for WCAG 3.0 conformance to address or ameliorate
the challenges described in this document?
PK: what are the approaches
JF: had some concerns in previous format, majority were
addressed, what would like to see. concern outlines why doesn’t
work.
... wrong hands would make look like can’t meet today
... need to demonstrate solutions are avaialble
... concerned about timing that would live in TR rather than
our own wiki for example
<AndySomers> COMMENT: The FAA has a Human Factors resource, I
am looking through it here for ideas on the stumbling blocks we
are discussing. Here's a link:
[10]https://www.hf.faa.gov/webtraining/Intro/Intro1.htm?tabid=3
49
[10] https://www.hf.faa.gov/webtraining/Intro/Intro1.htm?tabid=349
<sajkaj> Not rec track -- Note Track
<david-macdonad> is this the latest draft?
<david-macdonad>
[11]https://raw.githack.com/w3c/wcag/conformance-challenges-red
esign2/conformance-challenges/index.html
[11] https://raw.githack.com/w3c/wcag/conformance-challenges-redesign2/conformance-challenges/index.html
<Rachael> scribe: Rachael
Peter: To be clear, this document isn't saying that WCAG is
broken. Its talking about places where the conformance model is
difficult to practice.
John: When you get into a court of law, that is a real use
case, a real scenario. We all know this. You have collected in
one place. In the hands of the uninformed, it could pose a
problem.
Peter: I hear you. This is not an indictment of WCAG. It is
providing detail of something well known in the community -
that perfection isn't achievable. To argue against recognizing
the fact is concerning to me.
John: Its the presentation not the content that I have concerns
about.
david-macdonad: Is this the latest draft? It seems several
versions back.
... I had addressed some of those concerns and I was reasonably
satisfied but maybe this version is an earlier version.
Peter: The version that ends in conformance challenges redesign
2?
<JF> For clarity: my (our) concern is not about the content,
but a) how it was being "presented" to the uninformed public,
and b) the lack of any potential solutions.
Peter: I know this doesn't have the authors list trimming yet.
... but it did change the acknowledgements section and does
have what we discussed. When we talked, there was an
outstanding question that had been raised. We hadn't heard back
but can reopen that question to address those concerns before
the next survey. I believe you have the action.
Jeanne: Can I interrupt? We have moved off topic. What feedback
can we ask for from this?
Lucy: In academia, when we ask for something that has
challenges in it, we ask for a roadmap instead of a challenges
document
Peter: I am trying to process that.
Lucy: You've done a good job capturing the challenges but we
should reformat it as a roadmap. What are the priorities and
how can we research these and address them?
Peter: We have two questions. One focused on model and one
focused on content
<CharlesHall> @ Peter / Janina, the C.1 acknowledgement section
has an incorrect conflation of attribution, Charles Hall
(Oracle). we are Charles Adams (Oracle) or Charles Hall (MRM)
Peter: for example one might be to insist all videos are
captioned before we accept them. What are strategies for an
author trying to create an accessible website to fix issues
outlined?
Lucy: I think we need to look at this as what would a grad
student looking for their next thesis be asking?
Jeanne: When we put out a document, we ask for feedback. Peter
has come to us asking what questions does Silver have after
reviewing this document. I have thought of 2 but can't remember
them. Will come back. What Lucy said about priorities. What do
you think the priorities are for the challenges?
Peter: Can you expand on that?
Jeanne: Given the challenges we have identified, what is the
priority that the reviewers think we should address them in?
... if we can't get to all of them in Silver, what are the most
important?
... WCAG 3.0 not Silver.
... We could ask if there are specific success criteria in the
appendix that should be significantly revised and what
priority?
Peter: That is a great one.
Sean: We are at time.
<KimD> +1
<CharlesHall> +1
<AndySomers> Hear hear
<AndySomers> +1
<Jennie_Delisi> +1 for chairs!
Janina: Cheers to our chairs for running this so well
+1
s/sean/shawn
<kirkwood> thanks for taking over scribing Racheal!
<kirkwood> ahem.. Racheal/Rachael
[End of minutes]
Received on Wednesday, 11 March 2020 02:40:59 UTC