Re: Comments on WCAG 2.1 First Public Working Draft 28 February 2017

Hi Jan,

Many thanks for the input! Appreciated.

As there are a lot of comments relating to individual SC, it would be 
great if you could add these to the related individual Github issues. [1]

Thanks

Josh

[1] https://github.com/w3c/wcag21/issues

> Jan Richards <mailto:jan.richards2@gmail.com>
> 26 March 2017 at 05:03
>
> Hi all,
>
>
> Sorry that I haven’t been as active as I was hoping to be over the 
> past year. The new SCs are a welcome step towards a refreshed WCAG.
>
>
> Commenting on: https://www.w3.org/TR/2017/WD-WCAG21-20170228/
>
> *1.2.3 Audio Description or Media Alternative (Prerecorded)*
>
> I know this SC hasn’t changed, but I think it would greatly improve 
> the SC to bring this note out of the Understanding document and into 
> the normative guidelines:
> “Note 1: For 1.2.3, 1.2.5, and 1.2.7, if all of the information in the 
> video track is already provided in the audio track, no audio 
> description is necessary.”**
>
> **
>
> *1.3.4 Support Personalization (minimum)*
>
> This SC needs a lot more work before it’s acceptable as Level A or 
> even AA. It’s broad and ill-defined. What if context and important 
> info is within text prose? What kind of markup would be used? This one 
> needs a lot more thought..
>
> *1.4.3 Contrast (Minimum)*
>
> I know this SC hasn’t changed, but it would be nice to add an 
> exclusion for placeholder text if (and only if) there is a visual 
> label as well? Sometimes placeholder text is so high contrast it 
> tricks the user into thinking text has been entered.
>
> *1.4.10 Linearization*
>
> The exception is pretty huge here. Maybe the exception could be 
> avoided by just limiting the this SC to situations where the author 
> has flowed the body text into two or more columns of text?
>
> *1.4.11 Resize content*
>
> - Maybe be more clear that you mean horizontal scrolling (and then 
> perhaps clarify in the definition that this could be vertical 
> scrolling in differently flowing languages)
>
> - the “use” exception is going to be a big loophole here and needs 
> more work. How can a tester tell where spatial layout is necessary to 
> use and meaning?
>
> *1.4.12 Graphics Contrast*
>
> - Should there be a photograph exception here?
>
> - It’s odd that “essential” appears both in the SC wording and in the 
> list of exceptions.
>
> - I think the “thicker” exception might be better if it clearly 
> applied to lines, dots etc. since “graphical objects” can mean entire 
> image files.
>
> *1.4.13 Printing*
>
> - What can’t be printed? Video-only content?
>
> - Does it need any kind of note re: user agent constraints? (e.g. 
> assuming user agents can print these characteristics)
>
> *1.4.14 User Interface Component Contrast (Minimum)*
>
> This should only apply if the author has taken control of UI rendering 
> colours. It doesn’t seem right that if an author just relies on a 
> browser’s default rendering (e.g. of <button>) and the browser chooses 
> a low contrast rendering for its own reasons, that the author should 
> be penalized. Browsers have a responsibility here.
>
> *1.4.15 Adapting Text*
>
> - Verdana seems very specific. Does that internationalize well?
>
> - Does this SC rule out icon fonts?
>
> *1.4.16 Popup Interference*
>
> Does the Turn-off condition mean it has to be turned off before it 
> shows or is it ok to have a close button on the popup?
>
> *2.1.4 Speech Input*
>
> What are “the commands”? If I have a form with a bunch of fields and a 
> save button, is Save the command? Are the browser’s copy/paste 
> functions commands?
>
> *2.2.6 Timeouts*
>
> - I think this works for “tasks” with a start and end, like applying 
> for a job but not for more general sessions where a task is not 
> underway (e.g. signing into my bank account home page).
>
> - One week seems long. Is failing a site that keeps it for 5 days 
> reasonable?
>
> *2.2.7 Animation from interactions*
>
> I would be clear that you mean the entire screen, because there can be 
> small nested viewports etc.
>
> *2.2.8 Interruptions (minimum)*
>
> I agree re: Changes of Context, but interruptions needs to be defined. 
> Is a new email an interruption? A new airport gate assignment? May UIs 
> are like this.
>
> *2.4.11 Single-key Shortcuts*
>
> OK. I assume the mechanism must be in the content, not the browser or OS.
>
> *2.5.1 Target Size*
>
> I’m uncomfortable that these sizes are larger than the default 
> rendering sizes for elements like <button> in popular browsers. I 
> think WCAG should make it easy to use default element rendering (and 
> working on browsers to make sure they render well) and come down 
> harder when authors change default behaviour to reduce accessibility.
>
> *2.5.2 Pointer inputs with additional sensors*
>
> Should there be a note that the functionality might be in some 
> alternative way and that we aren’t forbidding these new methods? E.g. 
> to delete an email, I might press hard and slide left OR tap to select 
> it and then press a trash icon.
>
> *2.5.3 Touch with Assistive Technology*
>
> OK. Could it seem less imposing if it clearly applied to more complex 
> touch gestures?
>
> *2.5.4 Pointer Gestures*
>
> Some terms need definitions here. E.g. “Simple pointer gestures” 
> because a 3-finger “L” is simpler than a 4 finger “L”.  Maybe we 
> should set a standard? Max. 1 finger swipe?
>
> *2.6.1 Device Sensors*
>
> “is essential for the function and not using it would invalidate the 
> activity.” Seems like oddly repetitive phrasing. If something is 
> essential, then by definition it must be there (like a heartbeat 
> sensor for a heartbeat app).
>
> *2.6.2 Orientation*
>
> OK
>
> *3.1.7 Plain Language (Minimum)*
>
> I’m really not sure about this for A or even AA. This would be really 
> complicated to test and monitor. And how would anyone know what these 
> are? (“Use the most common 1500 words or phrases or, provide words, 
> phrases or abbreviations that are the most-common form to refer to the 
> concept.”).
>
> *3.1.8 Manageable Blocks*
>
> So “Select your language and press OK.” is not allowed due to it being 
> two instructions? Again, I’m not sure about this, though I think it’s 
> actually more feasible than 3.1.7.
>
> *3.1.9 Extra Symbols*
>
> Interesting idea. But does that mean that a police site could be WCAG 
> non-conformant because it listed it’s phone number without a symbol in 
> front of it? Which symbol? A phone? A police badge?
>
> *3.2.6 Accidental Activation*
>
> Quite a lot of conditions to parse for a relatively simple concept. 
> This wording (“either explicitly or implicitly as a platform's generic 
> activation/click event) seems to be a requirement on platforms (e.g. 
> Android, iOS) rather than page authors. Is that right?
>
> *3.2.7 Familiar Design (Minimum)*
>
> This needs more thought. It seems to prevent creative new UI patterns. 
> Would it have outlawed GUIs before they started?
>
> *3.2.8 Change of Content*
>
> Re: “The user has been advised of the behavior before using the 
> component.“: What component? None is mentioned prior.
>
> *3.3.7 Minimize user errors*
>
> I note “reliably” isn’t yet defined. It’s kind of important. Also, 
> will the user be notified? Will they be given an opportunity to correct?
>
> *3.3.8 Undo*
>
> The second bullet is a problem. What if I want to undo an action I 
> made two steps ago while keeping what I just did. That is impossible 
> in most Undo systems but seems to be required by this bullet.
>
> *3.3.9 Provide Support*
>
> I get the spirit, but I think people would likely point to things you 
> don’t want. E.g. to help understand a long document, users can read 
> the heading or the first paragraph etc.
>
> *_Definitions:_*
>
> *additional pointer sensor information *
>
> OK
>
> *
> *
>
> *author settable properties *
>
> This doesn’t make sense to me. Seems like it should mean all attribute 
> values.**
>
> *
> *
>
> *cardinal direction *
>
> OK
>
> *
> *
>
> *changes of content*
>
> Maybe changes made after it has been “initially rendered by the user 
> agent”. Should playing audio, video and animations be excluded?
>
> *
> *
>
> *common input error *
>
> Sorry but a weak definition. Basically, means an ever growing list of 
> things we tell you.
>
> *
> *
>
> *complex information*
> Needs work. I can easily think of things that meet the definition and 
> are not complex and things that don’t meet it but are.
>
> *
> *
>
> *concrete wording
> *Needs some work. The examples are “smoke, mist, a shout”. What about 
> “You must tell the truth in your answers”?
>
> *
> *
>
> *contextual information*
> Semantics is the biggie here. Virtually everything impacts meaning.
>
> *
> *
>
> *critical features*
> OK. How is this related to “Essential”?
>
> *
> *
>
> *critical service*
> This really talks about real-world impacts. I wonder if the term needs 
> to highlight that?
>
> *
> *
>
> *device sensor
> *OK**
>
> *
> *
>
> *CSS pixel*
> …at arms length according to the CSS3 doc. I think more of the CSS 
> definition should be included incuding the diagram (with appropriate 
> alt etc.)
>
> *
> *
>
> *easily available (or easily available mode or setting)
> *ok**
>
> *
> *
>
> *essential*
> I don’t think the term does want it needs to for all of the places 
> “essential” appears (e.g. 1.4.12 Graphics Contrast).
>
> *
> *
>
> *graphical object*
> So each letter? Not sure this fully works.
>
> *
> *
>
> *help*
> ok
>
> *
> *
>
> *identified context*
> Wow. Feasibility?
>
> *
> *
>
> *immediate surrounding color*
> ok
>
> *
> *
>
> *important information*
> Testable?
>
> *
> *
>
> *informational content*
> ok
>
> *
> *
>
> *long document
> *300 words?**
>
> *
> *
>
> *non-literal language*
> But this isn’t really the flip-side of “concrete wording”. Should it be?
>
> *
> *
>
> *non-numerical value
> *ok**
>
> *
> *
>
> *orientation*
> ok
>
> *
> *
>
> *personalization*
> Not just driven, but set up, rendered, etc.
>
> *
> *
>
> *platform assistive technology that remaps touch gestures
> *ok**
>
> *
> *
>
> *pointer input*
> ok
>
> *
> *
>
> *programmatic notification*
> ok
>
> *
> *
>
> *relative pronoun*
> ok
>
> *
> *
>
> *reliable and achievable (known to be)
> *empty**
>
> *
> *
>
> *significant animation
> *Be clear that viewport means screen, not just small sub-viewport.**
>
> *
> *
>
> *single-character shortcut
> *ok**
>
> *
> *
>
> *standardized technique
> *ok**
>
> *
> *
>
> *target*
> ok
>
> *
> *
>
> *understandable language*
> Are these identified?
>
>
>
> Cheers,
>
> Jan
>

-- 
Joshue O Connor
Director | InterAccess.ie