Silver Task Force and Community Group

30 Oct 2020


bruce_bailey, Lauriat, Francis_Storr, sajkaj, jeanne, SuzanneTaylor, Rachael, AngelaAccessForAll
Makoto, JakeAbma, ToddLibby, WilcoFiers, PeterKorn, CharlesHall, JanMcSorley, JenniferChadwick
Shawn, jeanne


<scribe> scribe: bruce_bailey

Presentation on Testing

Francis Storr will present on testing validation

FS has some model pages.

scoring test with accessible escape room sample site

FS: slide deck is not online

Rachael had created pages for other sample uses

Site is on GitHub repo, need to serve locally for captions

FaS is teesitng against alternative version.

For testing, paths are learn what is an accessible escape room

<Rachael> Actually it is a real site. I just happen to adminit and got permission for use from the others responsible.

Site includes NASA video just for testing.

Path 1 includes text alternative for images of text, and for testing had critical error on register link, so that was a critical error

Common clear words, rates go from not applicable to rating 4 of score of 1.7 or better.

scribe: scoring was too difficult, so I used 0, 1, 2 rating.
... of first 110 words, 28 are "less simple"

used xkcd simple writer for testing


Also excluded proper nouns and technical tersm, but still ended up with low rating

For Path 1, also scored captioned videos as 2 -- since NASA video had decent closed captions

scribe: On question that came up was use of term CAPCOM in captioning -- which seems to violate clear words

<sajkaj> It's a proper name!

<jeanne> That is an interesting question, because it needs to match what the video audio track. I don't think the clear words should apply.

scribe: Path 1, Heading Organzie Content, homepage was a DIV without rlevant roles
... but does not seem to interfer with completing task, so did not characterize this as critical error
... so fore scoring, 3 heading on page, divides page into sections appropriately, so scored that at 100%

FS: Path 1, VIsually Distinct Headings
... all heading were visually distinct, so 3 headings pass, 3 heading relevant, so again 100%
... Path 1, Heading Hierarchy
... H1 is dive wo relevant aria attributes
... other heading are H3 when should be H2, so scored 0
... Path 2 task is to find out if Escape Room event near me
... looked for airports near me, had altered page to use only airport codes
... Path 2, text alternatives for images of text
... no images on path, so NA
... also next test was NA
... Path 2, Use Clear Words
... scored rating of 0 because of obscure airport codes
... Path2: Heading Organize COntent
... Classified missing heading as Critical Error and a 0
... another score of 0
... Path 2: Heading Heirchary, another critical error, so another 0 score
... Path 3 was to learn more about the company
... first two fails for missing ALT were not on path, so NA
... Path 3: Visually Distinct Heading
... score of 4 headings passed when there were 5 headings, so 80% and a score of 4
... Path 3: Heading Hierarch
... 7 headings on path, 3 passed, 41% for a score of 1
... Plugged into excel document, noted right away that sheet is for views and I had paths
... Plugged scoreing results into four columns
... also plugged scores into functional categories and ended with 0.8 with 4 critical failures

Jeanne: Thanks FS for a ton of work, and as we anticipated, found a number of questions for application and exceptions
... definiately issues for next itteration. What are the exceptions? How do they interact?
... One thing I noted was that clear words should not apply to captioning -- since bigger need is for captions to capture what is spoken. So we might need to note an exception there.
... Big question is do you feel final score is reflective of accessibility.

Francis: 0.8 is definately a fail, but is 1 good enough.

Rachel: 4 is perfect

FS: It is a simple site, and I broke a number of things. So a 0.8 does seem about right.

Jeanne: Illustrates how much more guidence we need in the content for judging scenerios.

FS: The one thing I had to go back to is heading organizing content. I had to go back several times.
... that might just be getting use to the testing and scoring criteria.

Jannia: I wanted to get back to your question about missing H1 be a failure. I think I agree with you that the context is set from the home page, so description of failure should probably discriptive.
... I this were a long document or an e book, the missing H1 would really matter. For this site, and lots of other, a missing H1 on the home page is trivial and not a critical error.

<Zakim> Rachael, you wanted to say it might be good to pick an outcome and work through all the details.

FS: Thanks for affirming my thing. It works for a blog.

Rachelle: Question was about net step. So I think we need to through more examples.

FS: I had questions about spreadsheets I was supposed to use. I had two with slight different headings.

Rachelle: I want to suggest, you, me, Jeanne get together and straighten this all out.

Jeanne: We also need to detail the exceptions.

FS: Color contrast was also not clear to me, and in particular I did not know if that should apply to captions.
... I could also show testing for Deque site

<Francis_Storr> Text Alternatives For Images Of Text

<Francis_Storr> Text Alternatives For Functional Images

<Francis_Storr> Common Clear Words

<Francis_Storr> Headings Organize Content

<Francis_Storr> Using Visually Distinct Headings

<Francis_Storr> Heading Hierarchy

Jeanne reminds us that we also need to be concerned with Reliability, this shows we could/should provide more guidence.

FS: Yes, content off the path I did not score, but with the Deque site that was less obvious

<Zakim> Lauriat, you wanted to mention guideline/test-level and path-based scope of testing.

FS: Guidence for how much of a page view to include would be supper useful

S Lauriet: A few ways we could scope that, i think it should be at the test level

scribe: for this particular state, this would be the scope of testing
... for a state, the footers or whatever are really out of the current path
... but needs to be at test level, since with this next example, you describe the path is much more open ended

FS: Thanks, I have example of that.
... Deque shopping site for testing. Path 1, add laptop to your cart
... deck includes some screen shots
... pass / fail criter for text alternative for images of text
... for each there are six images, add to cart, and learn more. Critical buttons, right next to each other. Six buttons, but only Add To Cart is on critical path
... three critical errors of missing alt text, so scores 0
... Path 1, text alternative fun functional images: each of five images, w/o alt text, and sole content of links.
... score 0 due to critical errors w/ add to cart buttons
... Path 1: common clear words, Simple Writer fails thing like laptop and portable

Jeanne: The words probably would pass if 3000 words were used.

<Zakim> Lauriat, you wanted to mention plain language around technical topics.

Lauriat: Plain language does not prohibet using words that you are writing about, like laptop and keyboard, so does not prohibit technical termss

<Zakim> sajkaj, you wanted to ask about the limit of clear lang on tech

<Lauriat> +1 to sajkaj, I have no idea

Janina: But how far does that go? For a laptop, can I talk about SATA drives?

Jeanne: Thats okay, just need to have definitions, so including technical descriptions is okay because there is a plain language summary

FS: Example is with descripting creating silicon, could not have glossary, impossible to have clear words for some technical subjects

Jeanne: But would engineers object to plain language summary?

FS: Silicon design could have summary if blog post, but a technical document shared between colleagues, then plain language summary is a barrier to purpose of the communication

Jeanne: Given time, skip rest of presentation for now.

FS: Will figure out how to share decks.

Sub-group check-ins

Bruce thanks MichaelC for getting Myndex pulls incorporated

Susanne Taylor reports good progress with Errors sub group.

<jeanne> BB: XR met and looked at the different levels of what people mean by virtual reality. We had been looking at the most modern stuff, but there is a older level of XR that we didn't write about.

<jeanne> ... charles hall fleshed out the subset of XR needs from Functional Needs

Other groups absent, or did not report.

<sajkaj> https://www.w3.org/WAI/GL/task-forces/silver/wiki/Substantial_Conformance

Janina: Substantial Conformance has had two meeting
... working on purpose and questions for further discussing, scoping, strawman document for review
... we have participants new to working group, including web aim, so it is always nice to have new blood


Summary of Action Items

Summary of Resolutions

[End of minutes]

Minutes manually created (not a transcript), formatted by David Booth's scribe.perl version (CVS log)
$Date: 2020/10/30 19:05:11 $

Scribe.perl diagnostic output

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

Guessing input format: Irssi_ISO8601_Log_Text_Format (score 1.00)

Succeeded: s/no NA/so NA/
FAILED: s/patth/path/
Succeeded: s/Sample Writer fails thing like laptop/Simple Writer fails thing like laptop and portable/
Succeeded: s/Patth/Path/
Present: bruce_bailey Lauriat Francis_Storr sajkaj jeanne SuzanneTaylor Rachael AngelaAccessForAll
Regrets: Makoto JakeAbma ToddLibby WilcoFiers PeterKorn CharlesHall JanMcSorley JenniferChadwick
Found Scribe: bruce_bailey
Inferring ScribeNick: bruce_bailey

WARNING: No date found!  Assuming today.  (Hint: Specify
the W3C IRC log URL, and the date will be determined from that.)
Or specify the date like this:
<dbooth> Date: 12 Sep 2002

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]