W3C home > Mailing lists > Public > w3c-wai-ig@w3.org > July to September 2018

Re: User Stylesheets are Assistive Technology

From: Wilco Fiers <wilco.fiers@deque.com>
Date: Fri, 20 Jul 2018 00:40:35 +0200
Message-ID: <CAHVyjGM5TAdDz2BMSZFfx=EueY+nUKkF-zyZVmbH=cqhQ4w77w@mail.gmail.com>
To: Wayne Dick <wayneedick@gmail.com>
Cc: "Patrick H. Lauke" <redux@splintered.co.uk>, WAI Interest Group <w3c-wai-ig@w3.org>
Hey Wayne,
I really appreciate this e-mail thread. Going from its past work, AG hasn't
really seen custom stylesheets as assistive technologies. For example:
Color contrast is easily solved with a good custom style sheet, provided
the content author didn't do a few things that can get in the way of the
custom stylesheet. But rather than prohibit that, WCAG went the route where
there's some base level every site needs to meet, and whoever needs more is
just out of luck when it comes to those sites where custom styles don't
work.

The opposite was true for the WCAG 2.1 rule about word spacing (which is
where I can see your influence at work Wayne!), where the requirement
wasn't to have some minimum amount of word / line / paragraph spacing, but
rather to build the UI in such a way that it can be changed without
breaking. A very interesting move, and also completely the opposite
approach to the one taken with color contrast (or for reflow).

What it makes me wonder is some subset of "rules" can be defined for
building a site that support customisation. For instance, is it okay to put
informative images in CSS when there's an off screen text alternative?
Should CSS content ever be used to inject "content" in the WCAG meaning of
the word? It seems to me that, right now, there's no guidance on what you
should and shouldn't do with your CSS (and HTML?) in order to support good
custom stylesheets. But if there was, it might not take a CSS expert to
create them.

Wilco

On Thu, Jul 19, 2018 at 11:05 PM Wayne Dick <wayneedick@gmail.com> wrote:

> Nothing actionable today. That's why I am on the IG. I will formulate
> something actionable soon.
>
> I am gathering up ideas.
>
>
> Best Wayne
>
> On Thu, Jul 19, 2018 at 11:40 AM Patrick H. Lauke <redux@splintered.co.uk>
> wrote:
>
>> Is there anything actually actionable you're hoping to get from this
>> discussion? It's interesting, but I'm missing the context, or a concrete
>> "and for this reason I think we should do X".
>>
>> P
>>
>>
>> On 19/07/2018 18:19, Wayne Dick wrote:
>> > I will now focus on users with low vision.  It is a good example
>> because
>> > the scope is simpler than cognitive disabilities, but the solution
>> space
>> > is similar.
>> >
>> > The current model of AT does not work for people who have low vision
>> and
>> > cognitive disabilities. We need a personalized user interface. The
>> > access we need is like the access given by stylesheets when they work.
>> > We need selector level personalization.
>> >
>> > It is very clear that this cannot be provided by an AT that runs
>> outside
>> > of the browser. That would be an extreme breach of security. Right now
>> > CSS or browser extensions are the only way to achieve this result.
>> There
>> > are difficulties with both of these, but for now that is all there is.
>> >
>> > Ultimately there needs to be a way to pass style preferences to
>> browsers
>> > in a way that uses can get their visual style changes. Until then, CSS
>> > and extensions are it.
>> >
>> > Don't discount a tool that serves subject matter experts with
>> > disabilities. We do need to work. Ordinary users with low vision cannot
>> > write CSS. Most people who are blind don't write screen readers but
>> they
>> > need them. (The NVDA staff is a cool exception)
>> >
>> > The bottom line is that CSS is one of the only languages that can
>> safely
>> > mitigate the accessibility needs of the small group people with low
>> > vision who are IT professionals. It keeps many us working.
>> >
>> > It is extremely scary to live in a world where a basic method of
>> > accommodation can be taken away without notice, because nobody
>> > understands the extreme value of these tool to our lives. CSS is one.
>> > Configurable UI tools are another.
>> >
>> > Jon's comment on Windows 10 really illustrates the problem. For Jon the
>> > change was good. For me it made my 13 inch laptop impractical to use.
>> > Personalization is really necessary, but mainstream users think of it
>> as
>> > a nice feature, not a necessity. That is why people with fully sight
>> and
>> > no print disability can talk so casually about using CSS as an AT. CSS
>> > is not ideal, but for many it's infinitely better than nothing, or
>> > screen magnification.
>> >
>> > User stylesheets written to modify visual access are Assistive
>> > Technology. "hardware and/or software that acts as auser agent
>> > <https://www.w3.org/TR/WCAG20/#useragentdef>, or along with a
>> mainstream
>> > user agent, to provide functionality to meet the requirements of users
>> > with disabilities that go beyond those offered by mainstream user
>> > agents"  WCAG 2.0
>> >
>> > Best Wayne
>> >
>> >
>> >
>> >
>> >
>> >
>> >
>> >
>> >
>> >
>> > On Wed, Jul 18, 2018 at 5:28 PM J. Albert Bowden
>> > <jalbertbowden@gmail.com <mailto:jalbertbowden@gmail.com>> wrote:
>> >
>> >     "I think a major issue with user stylesheets is that there are no
>> stable
>> >     CSS-APIs that you could work against."
>> >
>> >     selectors are about as stable as they come and incredibly effective.
>> >     a generic stylesheet may not beat specificity 100% of the time, but
>> >     that shouldn't discount it, by any means.
>> >     moreover, any style sheet added to the document is going to have to
>> >     be scripted in, and even more likely in javascript.
>> >     so since we are already using javascript, lets just find the styles
>> >     that are not winning the specificity wars and then rewrite the style
>> >     at a higher specificity.
>> >
>> >     we can also use javascript to address frailty/brittleness in
>> >     selectivity; offer a nav/modal that appears on activation. read the
>> >     dom, present page elements in nav/modal with toggles/options, etc.
>> >     there are already a ton of bookmarklets that do most of this, pieces
>> >     of this, etc.
>> >
>> >     i actually think bookmarklets are more ideal here for
>> >     cross-browse/rplatforms, most particularly in terms of maintenance;
>> >     however, then i think it becomes an issue of user adoption. not many
>> >     people know about bookmarklets.
>> >
>> >     maybe i'm missing something entirely? i am certainly not an a11y
>> expert.
>> >
>> >
>> >
>> >     On Wed, Jul 18, 2018 at 7:01 PM, Chaals Nevile <chaals@yandex.ru
>> >     <mailto:chaals@yandex.ru>> wrote:
>> >
>> >         On Wed, 18 Jul 2018 21:40:11 +0200, Tobias Bengfort
>> >         <tobias.bengfort@posteo.de <mailto:tobias.bengfort@posteo.de>>
>> >         wrote:
>> >
>> >             I think a major issue with user stylesheets is that there
>> >             are no stable
>> >             CSS-APIs that you could work against. A user-stylesheet is
>> >             basically a
>> >             monkey-patch that will break on a regular basis.
>> >
>> >             In order to get this working reliably we would have to
>> >             convince authors
>> >             to trat their CSS as a public interface and announce
>> >             breaking changes
>> >             early on. I am not sure this reasonable.
>> >
>> >
>> >         We would. But in a world of CSS preprocessors and so on, it is
>> >         possibly easier than it might seem.
>> >
>> >         cheers
>> >
>> >             tobias
>> >
>> >
>> >             On 18/07/18 01:50, Wayne Dick wrote:
>> >
>> >                 There are lots of people who claim to be accessibility
>> >                 experts who
>> >                 disregard the value of user stylesheets as a significant
>> >                 technology to
>> >                 mitigate problems of visual interface. Actually they
>> >                 work quite well.
>> >
>> >                 This technology is used primarily be people who are left
>> >                 out of the
>> >                 mainstream ATs. They are a way to change colors, ensure
>> >                 a personalized
>> >                 contrast ration, control column width and many other
>> things.
>> >
>> >                 I use Safari because the browser will host user
>> >                 stylesheets. It is too bad
>> >                 that other browsers decided to stop supporting this
>> >                 important assistive
>> >                 technology.
>> >
>> >                 I think the AG should at least recognize that this is a
>> >                 form of assistive
>> >                 technology that is available in a technology landscape
>> >                 that offers almost
>> >                 nothing useful for most people with low vision and
>> >                 cognitive disabilities.
>> >
>> >                 For those who want to tell me how wonderful screen
>> >                 magnifiers are if I just
>> >                 used them correctly, don't bother. I probably know how
>> >                 to use them better
>> >                 than you. For my needs, screen magnification scores
>> zero.
>> >
>> >                 Wayne Dick
>> >
>> >
>> >
>> >
>> >         --
>> >         Chaals: Charles (McCathie) Nevile    find more at
>> https://yandex.com
>> >         Using Opera's long-abandoned mail client:
>> http://www.opera.com/mail/
>> >         Is there really still nothing better?
>> >
>> >
>> >
>> >
>> >     --
>> >     J. Albert Bowden II
>> >
>> >     albert@bowdenweb.com <mailto:albert@bowdenweb.com>
>> >     jalbertbowden@gmail.com <mailto:jalbertbowden@gmail.com>
>> >     https://bowdenweb.com/ <http://bowdenweb.com/>
>> >
>>
>>
>> --
>> Patrick H. Lauke
>>
>> www.splintered.co.uk | https://github.com/patrickhlauke
>> http://flickr.com/photos/redux/ | http://redux.deviantart.com
>> twitter: @patrick_h_lauke | skype: patrick_h_lauke
>>
>>

-- 
*Wilco Fiers*
Senior Accessibility Engineer - Co-facilitator WCAG-ACT - Chair Auto-WCAG

deque_logo_180p.gif
(image/gif attachment: deque_logo_180p.gif)

Received on Thursday, 19 July 2018 22:41:16 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 20:37:20 UTC