[access-uk] Re: Firefox, Internet Explorer, and MSAA [was: Re: Re: Dominos Pizza website]

  • From: Benjamin Hawkes-Lewis <bhawkeslewis@xxxxxxxxxxxxxx>
  • To: access-uk@xxxxxxxxxxxxx
  • Date: Sat, 18 Aug 2007 18:50:05 +0100

Alasdair King wrote:
Firefox's standards-compliance is irrelevant:
Until we know exactly what Steve was having trouble with, I don't understand how we can dismiss any difference between Firefox and Internet Explorer as irrelevant to his problems.
As Ray notes in his reply, many sites only work in other browsers so long as they ignore the standards that developers should code to and instead copy Internet Explorer's proprietary extensions and bugs. That's not always desirable, and in any case difficult because Internet Explorer is an exceptionally complicated black box: 

http://en.wikipedia.org/wiki/Black_box
Many web developers are used to spending half their time building their site to work roughly according to spec in every other popular browser, and then half their time working out why Explorer is breaking it and layering hack upon hack until it sort of works. 


His screenreader will have been built to handle Internet Explorer's 
presentation of web
pages, and Firefox support has been added at a later date and isn't so
advanced. it's the ability of Steve's screenreader to use it that is the issue.
If the screen reader does not support Firefox effectively, Steve's problem may represent a bug in the screen reader not in Firefox's use of MSAA. But Steve wasn't complaining about the limitations of his screen reader; he was making a specific claim that Firefox made poor use of MSAA and even more specifically that it wasn't exposing some links or link text correctly to MSAA. That can't be verified with any screen reader; you need to use an MSAA inspector tool. I wasn't willing to assume outright that Steve had not verified that Firefox was the problem, not his screen reader, and (if so) that Firefox's use of MSAA was the problem, not its HTML parsing or JavaScript evaluation. So I asked how he knew. 


Firefox support will improve in time if users demand it and
its MSAA implementation doesn't change. Ideally, in terms of support,
its MSAA implementation would be identical to Internet Explorer's,
since this is the de facto standard and the reference implementation
of a web browser for screenreader developers.
I think it's confusing to describe Internet Explorer as a "de facto standard" when the reality is it's just a very popular browser (with roughly 85% of users): 

http://en.wikipedia.org/wiki/Standard

http://en.wikipedia.org/wiki/Usage_share_of_web_browsers
This is especially true given the typical usage of "web standards" to refer to written specifications.
I realise you were probably using the term loosely but I don't think it's correct to call Explorer a "reference implementation" either. Software that fails to implement the relevant specifications, has no public test suite, and is closed source seems radically inconsistent with what Wikipedia (for example) has to say about reference implementations: 

http://en.wikipedia.org/wiki/Reference_implementation
Likewise it seems inconsistent with genuine Microsoft reference implementations, such as the Project REAL Reference Implementation that includes documentation, test data, and sample code: 

http://urlx.org/microsoft.com/c9e4d
While it's important to try to ensure that most web content accessible in Explorer is also accessible in Firefox, I fundamentally disagree that Explorer's use of MSAA should be copied slavishly. IMHO Mozilla should be trying to make the most accessible application they can with whatever tools are available, rather than mimicking the lowest common denominator and repeating Microsoft's mistakes.
Firefox needs to provide consistent web access across platforms, and Explorer's use of MSAA is not especially relevant to assistive technology that uses more advanced frameworks, like VoiceOver and Orca, or interacts with Firefox's DOM directly, like Fire Vox. Nor is Explorer's use of MSAA relevant to the various web features that Firefox supports but IE does not.
In practice, on the Windows platform, Mozilla have aimed to copy Explorer's use of MSAA, but with necessary additions to provide better standards support and accessibility to modern web technology that Explorer does not support (such as ARIA and MathML). In conjunction with other application developers (such as Sun and IBM) and assistive technology developers (such as GW-Micro and Freedom Scientific), Mozilla have been working to standardize their extensions to MSAA so that other developers can hack better accessibility onto MSAA in the same way: 

http://www.linux-foundation.org/en/Accessibility/IAccessible2
Some developers (NVDA's developers among them) seem to be already prioritizing Firefox over Internet Explorer, precisely because accessibility with Firefox is easier to achieve and provides greater benefits to end users. 


I use Firefox myself, but I'm sighted. Much better browser for me.
I don't want to see Mozilla blamed when it's not their fault, but I wasn't trying to start a discussion about which browser is "better" in general terms. SuI want to make sure that if there are bugs in browsers, screen readers, or important websites, they get fixed. Waiting around for browser developers or assistive technology vendors or web developers to happen to fix the bugs accidentally isn't as effective as identifying the bugs and reporting them to the people who can fix them. That's why it's important to press for details when problems are mentioned. It may even be that the problems turn out to be user error and not bugs at all. 

--
Benjamin Hawkes-Lewis
** To leave the list, click on the immediately-following link:-
** [mailto:access-uk-request@xxxxxxxxxxxxx?subject=unsubscribe]
** If this link doesn't work then send a message to:
** access-uk-request@xxxxxxxxxxxxx
** and in the Subject line type
** unsubscribe
** For other list commands such as vacation mode, click on the
** immediately-following link:-
** [mailto:access-uk-request@xxxxxxxxxxxxx?subject=faq]
** or send a message, to
** access-uk-request@xxxxxxxxxxxxx with the Subject:- faq

Other related posts:

  1. Home
  2. access-uk
  3. Archive
  4. 08-2007