Meeting minutes

Chair: Wilco
Present: Frank, Jemma, Wilco, Lori, Romain, Rob, Emma, Gian
Scribe: Jemma, Romain

[16:09] <Wilco> https://github.com/auto-wcag/auto-wcag/pull/11
[16:11] <jemma> any comment for the pull request?
[16:12] <Wilco>
https://github.com/auto-wcag/auto-wcag/pull/11/files#diff-f3bc3964bb5845e0c32f38c4a2fc9123R149
[16:12] <rdeltour>
https://github.com/WilcoFiers/auto-wcag/blob/08fb5525d4cb6ffe748e9b9af08f677035312f7d/drafts/SC1-4-1-link-in-text-style.md#block-like-element
[16:13] <rdeltour> https://developer.mozilla.org/en-US/docs/Web/CSS/display
[16:14] <jemma> romain:isn't it too simplistic in comparision to mozilla
one? there are many display properties.
[16:15] <jemma> rdeltour:complexity of 'inline"
[16:16] <jemma> wilco:could you leave comment to this pull? I will have a
look at it.
[16:16] <Frank> https://github.com/auto-wcag/auto-wcag/pull/11
[16:19] <jemma> frank: redundancy for link, a href and role =link?
[16:20] <jemma> wilco: it seems to be fine for this
[16:21] <jemma> rob:gradient discussion?
[16:21] <jemma> wilco; can you leave comments?
[16:21] <emmajpr> Hi Rob
[16:22] <emmajpr> Thanks shadi ... I keep forgetting the hyphen
[16:22] <jemma> romain:Distinguishing Border? should we look at contrast
issue too?
[16:24] <jemma> rob: consideration for transparency for filter, shadow..?
[16:24] <jemma> emma: thinking about stroke, ..to indicate headings
..something has significant shadow on it but it is not background.
[16:25] <jemma> wilco: that make things more complicated...;-)
[16:29] * rdeltour sorry I hadn't noticed the assumption :-)
[16:30] <jemma> emma: border and underline are separately rendered so would
it affect that?
[16:30] <jemma> emma: talking about pull request 11
[16:31] <jemma> wilco: I think it is not separate.
[16:32] <RobF> There are all sorts of assumptions that go into these rules
and how do we decide which are reasonable ones?  For instance, link text
could be distinguished using properties like stroke transparency, or other
filter effects.
[16:32] <RobF> Wilco's take was reasonableness is the standard.
[16:35] <jemma> robf: do we go about html, svg about rules?
[16:35] <jemma> robf: we also don't think about pdf?
[16:36] <jemma> wilco:those are will be separate rules.

[16:36] <Wilco> Topic: color contrast rule

[16:38] <jemma> wilco: so we can clearly explain whether rules will applied
to others such as svg or pdf..
[16:39] <jemma> wilco: state exclusion for svg and others
[16:41] <jemma> rob: ie. text top on the image, algorithmatically can
decide pass and fail but it would still require human decision
[16:42] <Frank> http://www.brandwood.com/a11y/
[16:42] <jemma> wilco: what do you think, Frank?
[16:42] <jemma> frank: it is complex
[16:45] <jemma> rob: if you do compare pixel by pixel, if text is not anti
alias, outside of pixel will be outside of image, ...target for color
contrast will be tricky.
[16:46] <jemma> wilco: may be we focus simple things first and then go
about technical detail and complexity.
[16:47] <jemma> emma: explain the way to evaluate color contrast from
Frank's example page.
[16:48] <jemma> emma: our other questions are also about 1)how much detail
we should go 2) how about text in videos and other format...
[16:49] <jemma> s/explain/emma explained
[16:50] <jemma> emma: the approach we are taking is finding text nodes and
find what colors are there and more. we are also trying to figure it out
how to process/evaluate those.

[16:51] <Wilco> topic: 3) Review: Form Division

[16:51] <Wilco> https://github.com/auto-wcag/auto-wcag/issues/16
[16:52] <jemma> wilco: does anyone take this?
[16:52] <jemma> s/take/would like to
[16:56] <jemma> wilco: I will update this to look at different form
elements based on accessible name calculation..ie. input field with address
-group
[16:57] <jemma> wilco: form field with duplicate names within the same
field will be a failure.
[16:57] <jemma> wilco: it is the similar logic like link should have a
distinguishable name

[16:59] <Wilco> topic: Review: Title

[16:59] <Wilco> https://github.com/auto-wcag/auto-wcag/issues/18
[17:00] <jemma> wilco: birkir was working on this. and he will rejoin the
group.
[17:01] <jemma> lori: what the definition of "title"? it is too broad here.
[17:01] <jemma> wilco: this is page title. I agree with Lori that it can be
broad.
[17:02] <jemma> emma: page title?
[17:02] <jemma> s/title?/ title will be better.
[17:03] <jemma> wilco: if there are two titles, it will show one title only
[17:04] <jemma> wilco: if there are two it is violation
[17:04] <jemma> emma: there is two page titles, it is problmatics
[17:05] <jemma> s/problmatics/problematic
[17:05] <jemma> emma: we had notion of comparing title with h1
[17:06] <jemma> emma: depending on correlation , we can see whether title
is ok or not
[17:08] <jemma> scribe:JaEunJemmaKu
[17:08] <jemma> chair:Wilco
[17:08] <Wilco> topic: Review: Pause, stop hide
[17:10] <rdeltour> scribenick: rdeltour
[17:10] <Wilco> https://github.com/auto-wcag/auto-wcag/issues/17
[17:10] <rdeltour> wilco: I think it's pretty good at this point
[17:11] <rdeltour> ... we should just pull out the blinking stuff in its
own rule
[17:11] <jemma> meeting: WCAG-auto
[17:11] <jemma> scribenick:jemma
[17:11] <rdeltour> emma: they all fall under 2.2.2
[17:12] <rdeltour> ... there s/b a link to the SC in the background section
[17:12] <jemma> present +Jemma
[17:12] <jemma> present +emma
[17:12] <rdeltour> wilco: right
[17:12] <jemma> present +rdeltour
[17:12] <rdeltour> emma: if they're all under 2.2.2, it makes sense to have
them all under the same rule
[17:12] <jemma> present +robf
[17:13] <rdeltour> wilco: Frank's comment: what if the blinking is removed
within a few seconds?
[17:13] <rdeltour> frank: I wouldn't do it myself, but someone could come
up with the idea
[17:14] <rdeltour> wilco: same thing is probably true for the blink element
[17:14] <rdeltour> ... what scenarios are these used in?
[17:14] <rdeltour> gian: someone using the blink element probably doesn't
know how to stop it with a timeout!
[17:14] <jemma> present+ jemma
[17:15] <jemma> present+ emma
[17:15] <rdeltour> all: [general agreement]
[17:15] <jemma> present+ rdeltour
[17:16] <rdeltour> gian: we can capture screenshots, wait a few seconds,
capture another one, etc several times, then compare the results
[17:16] <rdeltour> ... it's doable, it's been implemented
[17:17] <rdeltour> emma: that could be automated
[17:17] <rdeltour> wilco: I would say that it's something that warrants its
own rule
[17:18] <rdeltour> emma: gifs don't play and pause, everything is using
javascript to switch off images
[17:18] <rdeltour> wilco: you can't just look at the animation
[17:18] <rdeltour> emma: if JS has been added to allow a gif to not play
until someone clicks on it, there's a control in place
[17:19] <rdeltour> wilco: more comments ? [silence ensues]

[17:19] <Wilco> topic: aggregation

[17:19] <Wilco>
https://lists.w3.org/Archives/Public/public-auto-wcag/2016Sep/0009.html
[17:21] <rdeltour> wilco: it seems pretty similar to what the original
intent was, but a lot clearer. I  like it.
[17:22] <rdeltour> ... I'm wondering about the "if no relevant failure...
has been met" sentence
[17:22] <rdeltour> ... not sure about the relevance, or how that would work
[17:23] <rdeltour> frank: I think the idea is to first check the failure of
the SC, then successful implementation of the technique
[17:23] <rdeltour> wilco: this is a 2-step approach
[17:24] <rdeltour> frank: that would mean rules for any techniques and any
failure
[17:25] <rdeltour> wilco: I like the idea that it's in line with how WCAG
is structured, but not sold that there's a way to test that you've passed a
sufficient tecnique
[17:26] <rdeltour> ... I will email Alistair, see if we can get further
discussion on the mailing list.
[17:26] <rdeltour> emma: my understanding is it's right to think there's an
alternate way to do the auto-wcag thing? the expert system tool...
[17:27] <rdeltour> wilco: the expert system tool is a testing tool he
worked on. it's a method of testing. You posit statements and the expert
systems evaluate the statements.
[17:28] <rdeltour> wilco: any other thoughts on this?
[17:29] <rdeltour> lori: next month's meeting is on Thanksgiving
[17:30] <rdeltour> wilco: sorry I didn't realize that, I'll see if we can
move it
[17:31] <rdeltour> ... let's try to move it to Nov 17th


-- 
*Wilco Fiers* - Senior Accessibility Engineer