E-mail List Archives
Thread: Web applicationtesting
Number of posts in this thread: 9 (In chronological order)
From: sucharu
Date: Thu, Aug 04 2016 4:26AM
Subject: Web applicationtesting
No previous message | Next message →
Actually, I am looking for any online resources.
From: sucharu
Date: Thu, Aug 04 2016 4:34AM
Subject: Re: Web applicationtesting: ARIA wizards and controls
← Previous message | Next message →
Hi All,
Wish to learn, suppose you have choice to test on one browser, screen-reader combination on web page that has a wide application of ARIA controls and wizards, what would be your first choice?
Best,
Sucharu
From: Jamous, JP
Date: Thu, Aug 04 2016 6:55AM
Subject: Re: Web applicationtesting: ARIA wizards and controls
← Previous message | Next message →
I'd use NVDA as long as it supports the browser. NVDA is a standard screen reader and does not drill through HTML like JAWS. VoiceOver can be a shallow screen reader especially on iOS devices from my experience. In fact, I submitted a ticket to Apple about VoiceOver reading HTML incorrectly on iOS 9.3.2. For example, if the H1 is too big to fit on the iPhone screen, VoiceOver reports it as 2 separate H1 tags, when code-wise it is a single H1.
From: Jim Homme
Date: Thu, Aug 04 2016 7:10AM
Subject: Re: Web applicationtesting: ARIA wizards and controls
← Previous message | Next message →
Hi,
We use NVDA and Firefox here, but if a company's standard is, for example, JAWS and IE, we double check with that combination. As far as I am aware, NVDA depends the most on the accessibility information from the browser, and Firefox uses it the most. We think that gives us the purest test results.
Thanks.
Jim
=========Jim Homme,
Accessibility Consultant,
Bender HighTest Accessibility Team
Bender Consulting Services, Inc.,
412-787-8567,
= EMAIL ADDRESS REMOVED =
http://www.benderconsult.com/our%20services/hightest-accessible-technology-solutions
E+R=O
From: Sean Murphy
Date: Thu, Aug 04 2016 10:09PM
Subject: Re: Web applicationtesting
← Previous message | Next message →
Sorry, you should test NVDA and Jaws on both of those browsers.
NvDA has been enhanced to work better with Firefox then IE. Jaws works better with IE and does work extremely well with Firefox.
> On 1 Aug 2016, at 9:22 PM, Jamous, JP < = EMAIL ADDRESS REMOVED = > wrote:
>
> Sucharu,
>
> As suggested, but here is a more solid approach.
>
> JAWS with Internet Explorer
> NVDA with FireFox
> VoiceOver with Safari on Mac, iPad and iPhone
>
> If you cover the above, you can be in a great shape.
>
From: Sean Murphy
Date: Thu, Aug 04 2016 10:17PM
Subject: Re: Web applicationtesting: ARIA wizards and controls
← Previous message | Next message →
Single screen reader and single browser does not give the best result in your testing. Jaws and NVDA works different with each browser. Thus to get the best result, you should test both screen readers with both browsers. Chrome isn't well supported with screen reader's at this current time.
My work uses only Jaws with IE and Firefox for our testing. I do fall back on to NVDA when I want to validate issues to make sure it isn't Jaws or the browser.
Sean
As Jim stated, some companies focuses on a single screen reader like Jaws.
> On 4 Aug 2016, at 11:10 PM, Jim Homme < = EMAIL ADDRESS REMOVED = > wrote:
>
> Hi,
> We use NVDA and Firefox here, but if a company's standard is, for example, JAWS and IE, we double check with that combination. As far as I am aware, NVDA depends the most on the accessibility information from the browser, and Firefox uses it the most. We think that gives us the purest test results.
>
> Thanks.
>
> Jim
>
>
> =========> Jim Homme,
> Accessibility Consultant,
> Bender HighTest Accessibility Team
> Bender Consulting Services, Inc.,
> 412-787-8567,
> = EMAIL ADDRESS REMOVED =
> http://www.benderconsult.com/our%20services/hightest-accessible-technology-solutions
> E+R=O
>
>
From: sucharu
Date: Thu, Aug 04 2016 10:47PM
Subject: Re: Web applicationtesting: ARIA wizards and controls
← Previous message | Next message →
This is a link from WCAG 2.0 that includes information about User Agent Support for WAI-ARIA
http://www.w3.org/TR/WCAG20-TECHS/aria
As per this, NVDA support is partial.
So now, which is the optimum approach?
From: sucharu
Date: Thu, Aug 04 2016 11:05PM
Subject: Re: Web applicationtesting
← Previous message | Next message →
Hi,
As per my understanding, JAWS and NVDA behavior differently because of two facts
1. JAWS make some guess work in some situations like unlabeled forms
2. ARIA support is at different levels in both screen- readers
So, I think JAWS is likely to skip the errors easily catchable by freely available automated accessibility testing tools like AXE, a- inspector.
So, Is to test with automated tools first and then with JAWS, a better approach?
From: Jamous, JP
Date: Fri, Aug 05 2016 6:08AM
Subject: Re: Web applicationtesting
← Previous message | No next message
Automate than manual evaluate.
That's what I always tell our developers.
1. With automation you get to eliminate issues faster, but you miss out on details.
2. With a SR, you get first-hand experience with what is going on. Here it is important to use both JAWS and NVDA to identify the differences.
3. Once you identify the difference, you can write your code or modify it to find the balance or sweet spot.
I also advise our devs to refrain from using ARIA unless necessary. I have seen heavy ARIA sites that either crash JAWS or Internet Explorer.
ARIA was developed to be like medics. In other words, it comes to the rescue when there is an urgency to get something working. Otherwise, you should always check with your doctor to stay away from emergency situations. That's where you use proper semantic.
With proper semantic, all ATs function properly. With ARIA, each team or SR has its own way of coming to the rescue. I hope that makes sense.