WebAIM - Web Accessibility In Mind

E-mail List Archives

Re: Browser version advice in accessibility statement

for

From: Kevin Chao
Date: Nov 6, 2011 3:24PM

There's no excuse or reason at all one should not be running the
latest web browser and assistive technology, such as screen reader.
There's performance, security, accessibility, usability, support, etc.
benefits. Vendors/industry should not be limited due to users refusing
to update.

I think that HTML5 and ARIA should only be applied when needed, such
as rich, interactive, and dynamic sites. If a website is a fairly
static site and standard HTML4 markup will work, there's no need to
add additional complexity by adding in ARIA. However, if in the road
map of a company/site has plans on adding more rich, interactive, and
dynamic components; it's best to build in the HTML5 and ARIA support
during the R&D stage.

When it comes to web browser and assistive technology support, the web
is moving very quickly, and in order for companies/websites to keep
up; the minimum requirement and support must be very strict. It should
be forced to current and previous version, nothing more.

Firefox+NVDA, specifically 10 Nightly and 2011.3 Snapshot has the most
fastest, solid, accurate, and rich support for the web, specifically
HTML5 and ARIA.
NVDA works very well with Chrome 17 Canary and Google is actively
working on fixing a lot of accessibility support in many areas,
including HTML5 and ARIA.

Chrome+ChromeVox on Windows or Mac has very good support for the web
(rich, dynamic, and interactive), but ARIA support is a work in
progress.

Internet Explorer 9/8 does not have proper, robust, and rich
accessibility API's for ARIA and HTML5. Any AT/screen reader who
claims support for it is hacking support in, which results in poor
performance, instability, and buggy/inaccurate results.

iOS 5, VoiceOver, and Safari made a lot of great strides, especially
over iOS 4.3.X in improving HTML5 and ARIA accessibility support.

OS X Lion, Safari, and Voiceover just added ARIA support, it's quite
weak, and has a ways to go still. However, its support for HTML5 and
ARIA is far better than Window-Eyes, who has very weak web support,
and no HTML5 or ARIA.

I'm not sure how well SuperNova and System Access HTML5 and ARIA support is.

Opera web browser should not even been considered at all or mentioned
until it's fully accessible with various assistive technologies.

As a user myself, I don't think it's unreasonable at all to ask the
end-user of a website to upgrade to the latest web browser and AT to
fully utalize, leverage, and benefit from the Web 2.0 features of the
website. This should not be provided in a hidden link, but a visible
link, which when viewed provides the best combination, which on
Windows is Firefox+NVDA. NVDA, which is free and open source has the
best support, which results in the cost factor/excuse being thrown out
the window.

Most of the paid commercial screen readers are ones that I don't use
as their bloated, buggy, slow, and don't have great support for the
web, such as HTML5 and ARIA.

Kevin

On 11/6/11, Aaron Leventhal < <EMAIL REMOVED> > wrote:
> We definitely all need the compatibility list!
>
> Also, as far as inconsistent implementations of ARIA and how it affects
> older users ... If ARIA is implemented well the user should not have any
> trouble, because on almost l cases the accessibility behavior should be what
> they are already accustomed to. The inconsistency really points to the need
> for the ARIA developer portal where proper ARIA techniques are discussed and
> documented. We need aria testing tools, and we need a community that is more
> involved. We need to Identify and document the techniques and issues or
> inconsistencies in key implementations, and get after developers to fix
> their bugs.
>
> I believe this is all possible.
>
> Aaron
> -----Original Message-----
> From: "Birkir R. Gunnarsson" < <EMAIL REMOVED> >
> Sender: <EMAIL REMOVED>
> Date: Sun, 6 Nov 2011 03:36:27
> To: WebAIM Discussion List< <EMAIL REMOVED> >
> Reply-To: WebAIM Discussion List < <EMAIL REMOVED> >
> Subject: Re: [WebAIM] Browser version advice in accessibility statement
>
> Two minor thoughts on this (apart from me being outright impressed
> with the wealth of useful info on here, and discovering Henni's blog,
> which is excellent).
> 1. It seems to me that it might make sense for someone like webAIM, or
> a relatively objective and well respected third party to create and
> maintain a compatibility page for browsers and assistive technologies
> with regards to accessibility support (all the info is basically here
> in various blogs and in these discussions), and it might act as a
> disclaimer or source that people could link to in their own
> accessibility disclaimers. This would enable a somewhat generic
> accessibility statement and ensure a central page where changes can be
> made consistently as Assistive Technology is upgraded, or better info
> is available. Of course there could be issues, political or otherwise,
> why WebAIM per se, would not want to do this, but it seems to me like
> it would be a nice idea if some organization maintained a page with
> this info, allowing those who want to put browser and A.T. version
> support information on their websites (which I think is a great idea).
> If it is not centralized these statements will differ, be fragmented
> (not all people are aware of all screen readers, I see that
> Hal/Supernova support is absent in most lists for instance, something
> I could definitely help clear up), and out of synch, as it is not
> clear when each of these statements was updated and, hence, valid with
> regards to A.T. support for accessibility.
>
>
> 2. Perhaps I am somewhat alone in this opinion, I have certainly seen
> divided opinions on this. I think that when there is a free and open
> source screen reader out there, with navigation very similar to the
> Windows screen readers, that we can reasonably expect users who have
> older screen reading solutions, but require a more updated
> accessibility support, to go out there, download NVDA and learn how to
> use it. This is exaggerating the point certainly, especially when ARIA
> definitely is not as consistent in its implementation (for instance
> with keyboard support for flyout menus) as we'd wish it to be perhaps.
> There is a lot of users, who have recently lost their sight or are
> older and less computer savvy users, who simply can't deal with the
> complexities of ARIA, and are unlikely to frequent pages where ARIA is
> used aggressively. But, I think the availibility of quality ARIA
> support in a free and open source screen reader is a huge benefit, and
> can allow us to be a little more aggressive as developers utilizing
> these recent technologies to enable accessibility where other
> traditional means of A T support are not available. In short, whereas
> we need to preserve a balance, I wuld hate to avoid using ARIA because
> users of Jaws 7 can't utilize it. At this point, these users have the
> option to upgrade,and they need to.
>
> In my screen reader testing I have given a green light to features
> that NVDA 2011.2 supports, and also try to test two major versions
> back for the traditional major screen readers (now I test Jaws 12 and
> 13, with Jaws 11 testing being phased out).
> Perhaps this is overly aggressive, but there is definitely some
> responsibility for users to be up-to-date with their assistive
> technology, as long as it does not impose undue financial hardship on
> them.
>
>
> Cheers
> -B
>
> On 11/4/11, Aaron Leventhal < <EMAIL REMOVED> > wrote:
>> I agree Randy, although at least we can start to proceed conservatively
>> for
>> most content, using the fallbacks. And with the friendly message in place,
>> we can start to be aggressive with ARIA for the high tech crowd. We'll
>> find
>> out what's appropriate for the middle ground as we go -- and it will
>> change
>> (although not as fast as we'd like!)
>>
>> Aaron
>>
>> On Fri, Nov 4, 2011 at 4:27 PM, Randy Pope < <EMAIL REMOVED> > wrote:
>>
>>> HI Aaron,
>>>
>>> I'm in agreement with your thoughts. For many people with disabilities,
>>> updating or updating their assistive equipment and software pose a
>>> financial hardship. But again,,,it's very difficult to please everyone.
>>>
>>> With Warm Regards,
>>> Randy Pope
>>>
>>> -----Original Message-----
>>> From: <EMAIL REMOVED>
>>> [mailto: <EMAIL REMOVED> ] On Behalf Of Aaron
>>> Leventhal
>>> Sent: Friday, November 04, 2011 3:02 PM
>>> To: WebAIM Discussion List
>>> Subject: Re: [WebAIM] Browser version advice in accessibility statement
>>>
>>> Hi Christophe, I agree that every effort should be made to support older
>>> versions of JAWS and IE, and screen readers w/o ARIA support (like
>>> Windoew-Eyes). However, in dynamic content it's not always feasible. In
>>> many
>>> cases, depending on the audience and type of content, it simply makes
>>> sense
>>> to use ARIA.
>>>
>>> To make things better for users of JAWS 7 or IE7, etc., I advocate a few
>>> things:
>>> - Use "safe" ARIA techniques as a progressive enhancement when feasible
>>> - When "full" ARIA support is necessary, inform users with older
>>> technology
>>> of the need to upgrade and what their options are -- via the fancy
>>> automatic
>>> approach if we can get that to work well.
>>>
>>> As far as WCAG 2 compliance in that case, a strict reading of
>>> accessibility
>>> supported indicates we just need a free alternative -- such as Firefox +
>>> NVDA. Unfortunately as much as we'd all like to, we can't support five
>>> year
>>> old screen reading solutions in modern web content. We need to start
>>> educating users of the need to refresh their technology.
>>>
>>> Aaron
>>>
>>> On Fri, Nov 4, 2011 at 2:04 PM, Christophe Strobbe <
>>> <EMAIL REMOVED> > wrote:
>>>
>>> > Hi Aaron,
>>> >
>>> > At 16:54 4-11-2011, Aaron Leventhal wrote:
>>> >
>>> >> (...)
>>> >>
>>> >>
>>> >> Target audience:
>>> >> 1. If targeting the broad public (e.g. a government website), it
>>> >> seems necessary to stay on the safe side.
>>> >> 2. If targeting advanced technology users (e.g. a high tech company),
>>> >> it seems reasonable to use ARIA a lot more, and to require a more
>>> >> advanced browser - screen reader combo for content outside of the
>>> >> basics (documentation, support, etc.)
>>> >>
>>> >> Safe -- content for the broader public, or is primarily static HTML:
>>> >> . IE7+ with JAWS 7+, NVDA 2011.1+ or Window-Eyes 5.5+, Hal
>>> >> (version?), System Access (version?), etc.
>>> >> . Firefox 3.6 + with JAWS 7+, NVDA 2011.1+, Window-Eyes 5.5+ . Safari
>>> >> 4+ with VoiceOver on Snow Leopard or later . Mobile Safari and
>>> >> VoiceOver on iOS 4 or later
>>> >>
>>> >
>>> > This list reminds me of a similar list I wrote last year (in
>>> > deliverable
>>> > D3.1.2 for the AEGIS project; PDF at <http://tinyurl.com/6jjw8pz>;):
>>> >
>>> > * Internet Explorer 7 with JAWS 9 on Windows XP with Service Pack 3
>>> > (older versions of Internet Explorer and JAWS have no support for
>>> > WAI-ARIA),
>>> > * Firefox 3.0 with JAWS 9 on Windows XP with Service Pack 3 (Firefox 2
>>> > also supported MSAA and early drafts of WAI-ARIA),
>>> > * Firefox 3.0 with Window-Eyes 5.5 on Windows XP with Service Pack 3,
>>> > * Internet Explorer 7 with MAGic 11 on Windows XP with Service Pack 3,
>>> > * Internet Explorer 7 with ZoomText 9 on Windows XP with Service Pack
>>> > 3,
>>> > * Safari 3 on Mac OS X 10.5 with VoiceOver,
>>> > * Firefox 3.0 with Orca on Ubuntu 8.04 LTS ("Hardy Heron"),
>>> > * Firefox 3.0 with GNOME's built-in magnifiers on Ubuntu 8.04 LTS
>>> > ("Hardy Heron").
>>> >
>>> > This list was intended for creating accessibility support
>>> > documentation ("accessibility support" as defined by WCAG 2.0), hence
>>> > JAWS 9 instead of JAWS 7. (JAWS 7 is still in use, even in countries
>>> > with refund schemes for assistive technologies, e.g. Belgium.)
>>> >
>>> > Best regards,
>>> >
>>> > Christophe
>>> >
>>> >
>>> > Full -- content for a high tech audience or must be dynamic by its
>>> nature:
>>> >> . IE8+ with JAWS 10+ (unfortunately there is no live region support
>>> >> in
>>> >> NVDA+IE)
>>> >> . Firefox 3.6+ with JAWS 10+ or NVDA 2011.1+ . Safari 5+ with
>>> >> VoiceOver on Lion or later . Mobile browsers: to be determined
>>> >>
>>> >> (...)
>>> >>
>>> >> Thoughts?
>>> >>
>>> >> Aaron
>>> >>
>>> >> On Fri, Nov 4, 2011 at 10:33 AM, Kevin White < <EMAIL REMOVED> >
>>> >> wrote:
>>> >>
>>> >> > Hi All,
>>> >> >
>>> >> > I have a client who is doing some excellent work on creating an
>>> >> inclusive
>>> >> > and engaging website. In order to do so they are drawing on the
>>> >> > features provided in WAI-ARIA. This leads to some difficulties
>>> >> > regarding browser
>>> >> and
>>> >> > screen reader compatibility and we discussed how to address this.
>>> >> > My personal opinion is to use part of the accessibility statement
>>> >> > to
>>> >> highlight
>>> >> > the efforts but point out the need for users to upgrade but I was
>>> >> curious
>>> >> > to understand how people view this?
>>> >> >
>>> >> > My opinion is based on the idea that ARIA provides the opportunity
>>> >> > to
>>> >> help
>>> >> > users of assistive technologies but in order to do that there is a
>>> >> > need
>>> >> to
>>> >> > use a modern browser. User may not know this and by providing
>>> >> information
>>> >> > around this there is an opportunity to provide wider help.
>>> >> >
>>> >> > I would be interested to hear any other views on this,
>>> >> >
>>> >> > Thanks
>>> >> >
>>> >> > Kevin
>>> >>
>>> >
>>> >
>>> > --
>>> > Christophe Strobbe
>>> > K.U.Leuven - Dept. of Electrical Engineering - SCD Research Group on
>>> > Document Architectures Kasteelpark Arenberg 10 bus 2442
>>> > B-3001 Leuven-Heverlee
>>> > BELGIUM
>>> > tel: +32 16 32 85 51
>>> > http://www.docarch.be/
>>> > Twitter: @RabelaisA11y
>>> > ---
>>> > Open source for accessibility: results from the AEGIS project
>>> > www.aegis-project.eu
>>> > ---
>>> > Please don't invite me to Facebook, Quechup or other "social networks".
>>> > You may have agreed to their "privacy policy", but I haven't.
>>> >
>>> >
>>> >