<scribe> scribe: dmontalvo
<Wilco> clear agenda
<scribe> scribe: dmontalvo
Kathy: Elements with aria-hidden, Carlos worked out the last PR, I think it's approved
Wilco: Automatically playing audio and video
Will: Wasn't it approved last week?
Wilco: It did not yet get merged.
You were going to ask me for help but that did not happen
... This rule is proposed to be deprecated. Carlos is working
on combining several of these rules. We may not want to move it
forward to AGWG
... Should I mark this as deprecated even though it's still not
happened?
Kathy: Why are we deprecating this?
Wilco: This is a composite rule. We asked CG to merge these into one a while ago. Now Carlos is starting to work on this.
Kathy: The atomic rules won't be there any more?
Wilco: All three of this are
going to be deprecated, the two atomics and the composite, and
we'll just have one
... Will mark as deprecated
Daniel: Heading has non-empty accessible name. Will open issue with AGWG tomorrow
Wilco: I would not mind some reviews on these #1798
Helen, Will, and Daniel to review
Wilco: #1797. We should move this forward. I will assign myself
Trevor: I can take it
Jen: Yes please
Wilco: Will's and Helen's PRs can
be merged
... Karen has some work on the iframes PR
<Wilco> https://github.com/w3c/wcag-act/pull/526/files?diff=split&w=1
Kathy: There is a PR. We
requested everyone to take a look at it
... Input aspects were missing, they were called from several
rules
... These were video output, audio output, and source
code
... I see some people have put comments already
<Wilco> The byte data of a file from which a web browser or other user agent creates a page. For example, a browser may build up a web page from an HTML file, CSS file, and JavaScript file. The text of each of these three files is its source code. An ACT Rule could for example test for potential parser errors.
<Wilco> Source code is distinct from an HTTP response, which includes HTTP headers. It is the byte content of the file before parsing, which often results in an object model or syntax tree, or serialised versions of those. Notably in HTML, the outerHTML property of the root node can vary significantly from its source code. For the purpose of ACT rules, source files used in preprocessing such as PHP for creating HTML, or SASS for creating CSS are not cons[CUT]
Wilco: There is some edge cases we need to go to. It is not the https response, we should remover the php files
Trevor: The first sentences about
byte data, I am not sure what it means
... Second is about HTML, CSS, and JavaScript. The others could
be Applets and Java, which are bad examples
... Actually image byte maps might not be source code
<Wilco> The text content of a file from which a web browser or other user agent creates a page. For example, a browser may build up a web page from an HTML file, CSS file, and JavaScript file. The text of each of these three files is its source code. An ACT Rule could for example test for potential parser errors.
Wilco: Maybe we should say text content from a file ...
Trevor: Would "a browser user agent creates a page" -- is there anything beyond HTML, CSS, and JavaScript?
Wilco: PDF
Trevor: I could see that
... I fill that the last part of the second sentence leads to
think about further examples. Maybe a different example would
be helpful
Thom: PDF is not parsed by the browser, it's passed to a plugin
Wilco: Will review that part
Trevor: I'd be OK with us adding
an example that's less obvious
... I was trying to think of other technologies that get
rendered through the browser
... Others are pre processors that never touch the browser
Wilco: SVG, images, PDFs, you can open all of these in a browser
Trevor: I would be leaning
towards enumerating them all
... IF there's something that make us come back to this in the
future, we could add
Wilco: Could you put somewhere in there?
Trevor works on this live on the PR
More edits to the PR
Trevor: Browsers should be the same in mobile devices, except from some XML right?
Wilco: I am assuming we jumped
into this topic assuming everybody was on the same page with
this topic
... Each rules has input aspects
... We recently noticed that there are some of these which
where not defined, Kathy is working on these now
Trevor: It feels weird to me because the PDF is separate to everything else
Wilco: I don't think PDF has source code
Thom: The pDF has source
code
... the plugin parses it
Wilco: How about we just add SVG?
Trevor: That's fine
<Wilco> The text content of a file from which a web browser or other user agent creates a page. For example, a browser may build up a web page from an HTML file, CSS file, SVG file and JavaScript file. The text of each of these four files is its source code. An ACT Rule could for example test for potential parser errors.
<Wilco> Source code is distinct from an HTTP response, which includes HTTP headers. It is the text content of the file before parsing, which often results in an object model or syntax tree, or serialised versions of those. Notably in HTML, the outerHTML property of the root node can vary significantly from its source code. For the purpose of ACT rules, source files used in preprocessing such as PHP for creating HTML, or SASS for creating CSS are not cons[CUT]
Trevor: On both audio and visual
output, there is the condition that it has to be understood.
What does that mean?
... A recognizable language? Also captions?
Kathy: I saw "understood" was
used in language, so thought it was good to use this here
... Maybe it would be good to further explain here
... To me it means you have to comprehend what is said or what
happens, in order for captions to be accurate, audio needs to
be understood
Wilco: I would be a terrible reviewer of French captions
Will: What about parsed instead of understood?
Wilco: You have to know what the audio means.
Trevor: This is kind of subjective adn then further defined in the rules, right?
Wilco: Yes, your responsibility is to know what has been said or what the video is about. We don't get into how you get to know that
Trevor: I'm fine with that
Thom: Would it help saying "Is understood by the person or technology processing it"?
Wilco: "Understood by the tester"?
Kathy: Fine. I would suggest anything we do to clarify we use for language as well
Will: How about "comprehended by
the user"? Understand is confusing
... WE don't want to make it complex with long sentences
Wilco: I think I prefer understood
<JennC> Comprehended is probably more accurate, but understand is a bit more common knowledge
Daniel: What's the difference between understood and comprehended?
Will: Understood haas many meanings
Daniel: It may help if we restrict
Will: Maybe change it to "if you don't understand, you may not be able to complete the rules"
Wilco: I don't think we can make an assumption that it's going to be done by a person
Will: Just trying to move away from passive voice
Wilco: I've left a comment adding
the words "by the tester"
... We should update the language one as well
Helen: What about changed to
understandable?
... I use to use understandable as it points more to the
audience
<JennC> Understandable works
KAthy: If something is not understandable the caption will say something like "indiscernible"
<Wilco> Some rules can only operate on a Language aspect if the language is sufficiently understood by the tester, while other rules only require identifying the language. For example, a rule checking that an accessible name is descriptive can only function if the language is understood. A rule checking the correctness of a `lang` attribute requires knowing what language is used, but not the meaning of the words.
Wilco: Any objections to the changes we've been discussing?
RESOLUTION: Accept PR #526
RESOLUTION: Accept PR 526
Wilco: We have been asked by the
AG chairs to come up with more specific information about
updates to ACT Rules format 1.1
... AGWG is rechartering in October. We would need to have our
work listed very soon to be ready for the rechartering
... For the 1.0 version we have myself, Shadi, MaryJom and
Moe. Three of them are no longer working on this
... We discussed with Kathy and we asked Trevor, who said he
needs to check with his manager
Kathy: What's the deadline for providing the list to AGWG?
Wilco: End of month. Should discuss in our planning meeting next Tuesday
<Wilco> https://act-implementor.netlify.app/#/
Wilco: I have been working in a new solution for implementors
[Screen sharing]
Wilco: It's difficult to include
implementation reports by hand
... When things update you have to create a completely new
implementation
[Wilco shows screens of the new application: Start, New Implementation, Implementation Info]
Kathy: I like it
Wilco: I don't plan to integrate
this into the WAI website, I plan to have this as a standalone
tool
... I need to do more work to connect this to GitHub and other
things
Helen: Would there be export and import functionality?
Wilco: That's on my TODO
list
... Now that we have the redesigned rule pages, I am doing more
work on implementations
... Something I did now is to integrate IBM data
<Will_C> I gotta drop off and prep for a meeting in 5 minutes
Wilco: I have a PR open to integrate this. Tom, are you OK with this as-is? Do you want to take a look?
Thom: Does it update automatically?
Wilco: Yes, it rebuilds every
time the website re builds. IF you need anything updated, let
me know
... Now that i have these reports I can also include these
records into the WAI website
<thbrunet> Thanks!
<Wilco> https://github.com/w3c/wcag-act-rules/pull/80
<Wilco> https://deploy-preview-80--wai-wcag-act-rules.netlify.app/standards-guidelines/act/rules/
Wilco: I opened a PR for you to
see what I am proposing
... There is a new section on deprecated rules
... Do people need more time to review this?
Helen: I am happy
Jen: It's good
Wilco: I will automate this