<kirkwood> agree with JF on sharing
Jeanne: Michael Cooper thought we should start sharing the sketch of the conformance because we could use the feedback and help in figuring out the hard stuff.
Minutes from where we left off: https://www.w3.org/2018/11/30-silver-minutes.html
Shawn: How to reflect the an important image is broken.
<Lauriat> https://docs.google.com/document/d/18JyGF-AK8Qgq7DPyVlDYmxoj6814rORxuCf0l0oSb7U/edit
Method for Language of Page: https://docs.google.com/document/d/18JyGF-AK8Qgq7DPyVlDYmxoj6814rORxuCf0l0oSb7U/edit#heading=h.6aa9aegss0ly
Charles: Heuristic evaluations
often still use the Nielsen scale developed in the 90s. It is a
1-10 scale that is scored for each page, then the overall score
is tabluated and then recommendations for remediation are
given.
... in Silver, it could be "It is understandable" so you would
tell if every image is understandable and scale it as 1-10,
then say anything over 5 passes
... Shawn: So in the example of the hotel booking site, the
score for alternative test would score "with difficulty"
because the stars don't have explanatory alt text.
... The other testing and results of those tests are the
justification for the heuristic score, but the heuristic
evaluation is conducted independently of that test.
... run the test to make sure that alt text exists
<Lauriat> …Shawn: The heuristics would stay at a higher task level, with the applicable tests that apply given the technology and build up the justification of the heuristics score.
<scribe> scribenick: jeanne
Shawn: Someone who doesn't know
what tests apply, have to run through a list of tests and
select which tests are applicable.
... experienced people know the applicability of tests and go
from the content or DOM down
... it's forming a framework of how people should think about
heuristic evaluation and say that the heuristics may or may not
apply.
Charles: that's one way to approach it, what I'm suggesting is that we have a high level set of heuristics that is more like the Principles: Is it understandable, is it perceivable, etc.
Shawn: Having the heuristics at the task level -- not linked directly to the tests -- but give people guidance of "this is what you need to think about when you are assessing it".
Charles: A narrower set of
heuristics help with the cost of testing
... should the higher level heuristics be Principle based or
human modality based.
... should it be "everything interactive must be touchable".
One heuristic can be the modality of touch, one can be modality
of sight
... then we can account for an evaluation that has judged them
all.
... this can be against the entire site once all the other
modalities
Jeanne: I would strongly recommend that we align with the EN 301 549 and section 508 modalities, as people who conform to those standards already have to do that.
Shawn: I think 1-10 is too
granular. I think 1-4: Impossible, difficult, decent,
awesome
... I want to have a scale that includes awesome, is that
people keep asking for a scale where people can do better than
"fine"
JohnKirkwood: This is a problem
that I run across. In the courtroom, that's where the problem
with black-and-white accessible. A jury thinks: This is or is
not equivalent for this community.
... it's not successful, it's equivalent for different groups
of people.
Charles: if the heuristics are human modality based, and we have a defined set of modalities to test against, and the subject of the evaluations should be a "separate but equal" site. We will not have "text-only" or accommodation sites.
Jeanne: I agree with the sentiment, but there are problems with requiring that. For example, an organization who is being sued for lack of audio description in their courseware, but the platform doesn't support a video channel. The organization is making a separate set of videos with the audio description included in the main video channel.
JohnKirkwood: I agree.
John gave an example the scribe missed.
Charles: How do we ensure that the evaluation is conducted against the main site.
<kirkwood> for example some sites translate into 9 languages and legal precedent around it and ensure same experience. can look to as an example.
<kirkwood> One thing that i feel missing is navigation, and task completion
Jeanne: I think we still will need to have individual heuristic tests, especially for some of the cognitive disability needs.
JOhnKirkwood: The Principle missing in WCAG is Navigable. It is keeping people out of the internet world because they are having a hard time navigating it.
Charles: I think I have been asking myself: "Principles vs Modalities". I think the answer is both. There are principles but they are lacking some items.
JohnKirkwood: With AR and VR, the way Silver will be transformational in the future, it will be the ability to get through information and to know where one is in the information.
Jeanne: +1
<kirkwood> dounds good
Shawn: A good activity for Friday would be to sketch out an outline of Conformance and mark what is hand-wavy or has problems and go share it and ask for help.
Charles: I will put together this list of heuristics, but I don't know how to fit it into conformance
Jeanne: I will pull together the images and text from the different slide decks.
<kirkwood> sorry my audio dropped
Shawn: I started the document and will send out a link.
<scribe> ACTION: Charles to create a list of heuristics
<trackbot> Created ACTION-197 - Create a list of heuristics [on Charles Hall - due 2018-12-11].
<scribe> ACTION: Jeanne to add information from slide decks to the new Conformance document.
<trackbot> Created ACTION-198 - Add information from slide decks to the new conformance document. [on Jeanne F Spellman - due 2018-12-11].
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: Charles Lauriat kirkwood AngelaAccessForAll JF Makoto KimD Jeanne Jennison Regrets: Jennifer MikeCrabb Found ScribeNick: jeanne Inferring Scribes: jeanne Found Date: 04 Dec 2018 People with action items: charles jeanne 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]