- From: Sajka, Janina <sajkaj@amazon.com>
- Date: Fri, 15 Nov 2019 13:41:26 +0000
- To: Detlev Fischer <detlev.fischer@testkreis.de>, Silver Task Force <public-silver@w3.org>
- Message-ID: <4545d636a29745e683d2cf3f74962e47@EX13D28UWC001.ant.amazon.com>
Thank you, Detlev.
I’ve begun the process of moving your contribution into the Challenges document. There’s much more editing to do, but you can find your additions in this github branch:
https://raw.githack.com/w3c/wcag/conformance-challenges-working-draft_Detlev1/conformance-challenges/
Best,
Janina
From: Detlev Fischer <detlev.fischer@testkreis.de>
Sent: Thursday, November 14, 2019 2:12 PM
To: Silver Task Force <public-silver@w3.org>
Subject: Additional SCs that cannot be checked automatically (or only partly, incompletely)
Hi,
Here are some comments on a further 20 SCs that cannot (fully) be checked automatically in my opinion. Others may disagree, and advances in automatic checking alorithms may improve the situation.
best,
Detlev
-------------
1.4.13 Content on Hover or Focus
As content needs to be brought up by hovering the pointer over it (or keyboard-focus it) to then determine whether the 3 criteria of 1.4.13 are met, this is by necessity a human test.
2.1.4 Character Key Shortcuts
single key presses can be applied to content via a script but whether and what these keypresses trigger can only determined by additional human checks.
2.2.1 Timing Adjustable
There is no easy way to check automatically whether timing is adjustable. Ways of controlling differ both in naming, position and approach (including dialogs(pop-ups before the time-out) and depend on the way the server registers user interactions (e.g. for automatically extending the time-out).
2.2.2 Pause, Stop, Hide
Typically the requirement to control moving content id provided by some interactive controls placed in the vicinity of moving content, or rarely on a general level at the beginning of content. The fact that position and naming vary means that an assessment is mostly human task (it will involve checking that the function works as expected).
2.3.1 Three Flashes or Below Threshold
May be automatable but it involved an assessment of the area of flashing, so probably human. As this Failure hardly ever occurs, I just mention it for competeness.
2.4.1 Bypass Blocks
You can determine whether native elements or landmark roles are used but probably not if they are used to adequately structure content (are they missing out on sections that should be included). The same assessment would be needed when other Technique sar eused (structure by headings, skip links).
2.4.2 Page titled
Easy to check automatically whether the page has a title but the descriptive check is a human one. Matching title to h1 will not cover pass cases where title and h1 differ.
2.4.3 Focus Order
Focus handling with dynamic content (move content to coustom dialog, keep focus in dialog, return to trigger) do not look like they will be automatable any time soon.
2.4.4 Link Purpose
Automated can check fo rthe exisitance of same named links and probably check whether they are qualified programmatically, but checking whether the context adequately serves to describe link purpose seems like still involving himan judgment.
2.4.5 Multiple ways
Automatically checing the presence of several ways (e.g. nav and search) helps but may miss cases where exceptions hold (all pages can be reached from anywhere) so needs a human check? But maybe algorithms already exist to check that.
2.4.6 Headings and Labels
The determination whether headings are descriptive depend on an assessment of the context of web content headed or labelled and is therefore predominantly a human assessment.
2.5.1 Pointer Gestures
I am not aware of an outomated check that would detect complex gestures - even when a script indicatesthe presence of particular events like touchstart, the event called would need to be checked in human evaluation.
2.5.2 Pointer Cancellation
When mouse-down events are used (this can be done automatically), checking for one of the four options that make it OK looks definitely like a human task.
2.5.4 Motion Actuation
Event suspect may be detected automatically but whether there are equivalents for achieving the same thing via user itnerface components will require a human check.
3.2.1 On Focus
I suspect it is hard to check automatically whether a change caused by focussing should be considered a change of content or context.
3.2.1 On Input
I suspect it is hard to check automatically whether a change caused by inputting stuff should be considered a change of content or context, or to automatically detect whether relevant advice exist before the comoenent in question.
3.3.1 Error Identification
Whether the error message correctly identifies and describes the error will often involve human judgment.
3.3.2 Labels or Instructions
Edge cases (are labels close enough to a component to be perceived as a visible label) will need human check. Some label may be programmatically linked but hidden or far off. Whether instructuins are necessary and need to be provided will hinge of the content. Human check needed.
3.3.3 Error Suggestion
Whether a suggestion is helpful / correct will often involve human judgment.
4.1.2 Name, Role, Value
Incorrect use of ARIA constructs can be detected automatically but constructs that appear correct may still not work, and widgets that have NO Aria (but would need it to be understood) can go undetected. Human post-check of automatic checks seems necessary.
Am 14.11.2019 um 15:10 schrieb Sajka, Janina:
From: Detlev Fischer <detlev.fischer@testkreis.de><mailto:detlev.fischer@testkreis.de>
Sent: Thursday, November 14, 2019 7:16 AM
To: public-silver@w3.org<mailto:public-silver@w3.org>
Subject: Re: Minutes from Silver Conformance on 12 November
Hello, Detlev:
Your additions would be very welcome. I can take them in whatever way works best for you. Text in email, or an attachment in email, a github pull request, etc., -- they’re all good.
IMPORTANT: Be sure you’re adding to, or suggesting edits in the Working Draft which has all the latest edits and additions. The Working Draft will always be at this link:
https://raw.githack.com/w3c/wcag/conformance-challenges-working-draft/conformance-challenges/
I cannot promise I can get your text into the FPWD, because that draft needs to be locked down in a few hours. However, there will be additional drafts. I think we’re all agreed the document isn’t finished, and the idea has always been to get wide input so that we can have the best understanding possible of the problems with WCAG we need to try to do better with in the next version.
Best,
Janina
There are some WCAG 2.1 SCs not listed under challenge #1 (human involvement needed) which also need human involvement. Is the aim to provide comprehensive coverage here (i.e., add those SCs that are still missing) or is the list just indicative of the general problem? If other SCs should be added I am ready to provide draft text for (at least some) that are missing.
Detlev
Am 14.11.2019 um 01:20 schrieb Sajka, Janina:
Minutes from the Silver Community Group teleconference of Tuesday 12 November
are provided here.
Hypertext minutes are available at:
https://raw.githack.com/w3c/wcag/conformance-challenges-working-draft/conformance-challenges/
===========================================================
SUMMARY:
* Work on our Challenges with Conformance document has been moved to
github preparatory to its publication as an AGWG First Public Working
Draft (FPWD).
* Work on the conformance section of the Silver FPWD updated and
discussed.
* Also additions to User Scenarios.
===========================================================
W3C
- DRAFT -
Silver Community Group Teleconference
12 Nov 2019
Attendees
Present
jeanne, janina, PeterKorn, bruce_bailey
Regrets
Chair
jeanne
Scribe
janina
Contents
* Topics
1. update on Challenges
2. FPWD Conformance section
3. adding to user scenarios
* Summary of Action Items
* Summary of Resolutions
___________________________________________________________________________________________________________________________________________________________________________________________________________________
present=
update on Challenges
<jeanne> Janina: We have moved it to Github. Not much as changed except the technology of Github. Mostly because W3C publications come from Github
https://w3c.github.io/wcag/conformance-challenges/
<jeanne> ... I went through the document line by line comparing it with Google docs.
<jeanne> ... the Google doc is marked as deprecated and people are referred to Github\
https://github.com/w3c/wcag/issues?q=is%3Aissue+is%3Aopen+label%3A%22Challenges+with+Conformance%22
<bruce_bailey> good first impressions!
<jeanne> ... the margin notes were moved to issues.
<bruce_bailey> can we loose quote and slash or unquote in title ?
<jeanne> Janina: Is it ok with the group to combine some of the issues in Github? The issues can have back-tracing to the original posting.
<jeanne> ... there are instructions on using Issues in Github with email.
<jeanne> ... comments can be replies and email with changes, to the branch merge or pull request depending on the level of Github savvy.
<jeanne> ... these are issues with the Challenges rather than broader issues with WCAG.
<bruce_bailey> +1 to publishing challenges doc instead of Silver
Lost audio! brb
<jeanne> Peter: IN discussing it with the chairs, we agreed to publish it in AGWG. It pertains more to WCAG than to Silver.
<jeanne> Janina: We got a lot of good comments from Mary Jo, who isn't a Silver participant.
<jeanne> Jeanne: I'm glad to see this moving along. It will help more people understand the problems that Silver is trying to solve.
<jeanne> Peter: Timeline-wise - we are interested in comments this week before bringing it to a vote at the AGWG meeting.
<jeanne> Jeanne: Do you want to be able to say that it has the approval of the Silver Task Force.
<jeanne> Peter: I want as much additional review and comments before going to AGWG. Thumbs-up from Silver Task Force members can't hurt.
<jeanne> Bruce: This is the natural outgrowth of the Silver process. Even though the content people haven't been working on it, it came out of the Silver process and that's important to emphasize.
FPWD Conformance section
<scribe> scribe: janina
jeanne: Wanting to know what else people would like to see in it, or what we should remove?
peter: I think I sent you content? Yes? No?
[people look and decide not]
Peter: Seems I'm working on something for the Explainer
<jeanne> Janina: Explainer needs a what we mean by organizations
<jeanne> https://raw.githack.com/w3c/silver/ED-Draft-js-Oct/guidelines/explainer.html
<jeanne> Main FPWD draft https://raw.githack.com/w3c/silver/ED-Draft-js-Oct/guidelines/index.html#scoring-conformance
Argh. Lost audio again!
jeanne: In the process of moving architecture to Explainer because it fits better there
Peter: Just now sent you text
... Selecting a bullet to 3.3 to note we want to address some of the challenges we've found with 2.0 model
Joe: Asks about process practice
Jeanne: Notes we've found challenges with 2.0 for all sized orgs
Peter: Yes could be a separate bullet, or an addition to this bullet
Jeanne: Another important thing to include is to move away from the 100% binary fail or pass model
<jeanne> TOlerances
Jeanne: We are supposed to call that tolerances
... Will be running potential language by Shadi who has spent a lot of time on this
... One example Shadi raised is driving 1 mile over the speed limit would not likely draw a ticket, but 30 mph would
... Another Shadi example was that the same threshold bump in the hotel entry was more of a barrier for a wheelchair userfollowing a long international flight, than just arriving for dinner at the hotel after a cab
ride from home
... Shawn talks about a google problem where they cannot reach WCAG conformance because of a hidden iframe that no user would ever actually reach, but tools find it and fail the page
peter: shifting to tolerances is powerful and worth exploring futher
... the iframe is a great example--the user never gets there so it doesn't matter, but it fails
... Or an image with a caption that explains all that's needed
jeanne: has minimal impact
peter: doesn't material impede the user from getting done what they need to get done
... liking "materially impact" but can further consider
jeanne: So, if we were to putthis in our goals, what would we say?
peter: a conformance model that would tolerate errors that don't materially impact pwds
... Consider that a first draft!
jeanne: Everyone OK with "materially impact?"
[agreement]
jeanne: anything else?
... will continue to raise this
adding to user scenarios
<jeanne> https://docs.google.com/document/d/1PsMDJ4MLmTv8WeIhtFsGj-nJDwJhuzI8ATMAS2UuWa4/edit?ts=5db83f59#heading=h.ywxwuch4lj7t
jeanne: Notes Kim Dirks has added content
<PeterKorn> We're at the 45 minute mark, so I'm going to re-start my WebEx so I can keep talking (not that I'm trying to right now)
jeanne: Kim started in #5: user generated content
... i.e. how to encourage a11y in uploaded videos and photos
<jeanne> What about uploading photos, graphics and video? How to encourage or ensure accessibility? (with uploaded photos and videos, 1.2.1-1.2.9 and 1.1.1 are often violated.)
<jeanne> Third party content can be copyrighted or have other intellectual property restrictions, making it difficult or impossible to adapt that content and make it accessible.
<jeanne> Always Under Development: As a consequence of an evolving website, there are sometimes older legacy components which may be very difficult and expensive to update as new SCs are implemented.
<PeterKorn> sorry; it is again taking forever for WebEx to connect me.
jeanne: Anything to add for "highly dynamic content?"
Summary of Action Items
Summary of Resolutions
[End of minutes]
___________________________________________________________________________________________________________________________________________________________________________________________________________________
Present: jeanne janina PeterKorn bruce_bailey
Found Scribe: janina
Found Date: 12 Nov 2019
----------------------------------
Janina Sajka
Accessibility Standards Consultant
sajkaj@amazon.com<mailto:sajkaj@amazon.com>
--
Detlev Fischer
Testkreis
Werderstr. 34, 20144 Hamburg
Mobil +49 (0)157 57 57 57 45
http://www.testkreis.de
Beratung, Tests und Schulungen für barrierefreie Websites
--
Detlev Fischer
Testkreis
Werderstr. 34, 20144 Hamburg
Mobil +49 (0)157 57 57 57 45
http://www.testkreis.de
Beratung, Tests und Schulungen für barrierefreie Websites
Received on Friday, 15 November 2019 13:41:33 UTC