RE: 1.3.1 question

Hi Mike,

 

A great question, and a complex one to respond to. Let me try it this way:

 

We all have observed that the use of headings and landmarks are a boon to usability, and in that I don’t think anyone would argue to the contrary. 

 

But, just like automatic transmissions in automobiles, the absence there-of (i.e. a manual transmission) does not make the automobile non-functional – it may be *harder* to drive, and some younger drivers may not have any experience with a manual transmission (i.e. a usability problem), but the bottom line is that the automobile is NOT broken, it just doesn’t have an automatic transmission.

 

The same argument can be applied to legacy pages that are on the web today – some built long before aria and/or HTML5 landmarks were invented: they aren’t broken, they just don’t have [sic] automatic transmissions (if you catch my point).

 

> …increased use of headings and regions I worry that we may be underestimating their benefit

 

There is no doubt that using the most current technology in the way it was intended will offer maximum benefits – I mean, why else would the new technology emerge if not because of benefits? As developers, accessibility advocates and teachers, as folks who care about our users and offering the best usability to all of them, yes, sure, use the best techniques possible, whenever and however you can.

 

But we have to draw the line at a point where our desire to offer the best becomes an “undue burden”, a weasel-word concept in the law, sure, but one we cannot simply dismiss out of hand. Insisting that legacy content comply to a newer Best Practices benchmark can create an undue burden on content owners dealing with 100’s, 1,000’s and even 100,000’s of web documents. There is a real cost to maintaining that amount of data, and constantly having to return to tweak each conformant document every time a new Best Practices technique emerges 100% fits the definition of “undue burden” – if it conformed to WCAG 2.0 in 2009, it should remain conformant in 2016 (manual transmission and all…)

 

This is the difference between STANDARDS and BEST PRACTICES, and as developers and content creators, we should be striving for Best Practices, and not accepting anything less than the Standard. I don’t see a dichotomy there, it defines a range, with the Standard being the minimum, and the Best Practice being the maximum (and, if you hit a “Good Practice”? Not the Best, but better than the minimum? Yay still.)

 

As accessibility advocates, yes, our primary constituency is always the users first, but we must never forget that there are other constituents in the mix, and we owe them the respect and honesty from our craft as much as our end users. Honesty goes a long way, and I’ve found personally that selling the client on [sic] an automatic transmission (ha!) isn’t that hard, as long as I am prepared to admit that the same model comes with a standard transmission as well: costs a little less, but has significantly less take-up (“kids these days don’t drive stick”). I find that a far more honest approach than suggesting that all cars MUST have an automatic transmission because, you know, it’s 2016.

 

> Ultimately our objective has to be how to best serve the needs of users, however. 

 

And so, I will suggest that one of the ways we can best serve our users is to be honest with our clients. Don’t place artificial barriers in *their* way (“Every page MUST have aria-landmarks”), don’t try and use WCAG (or any other standard) as some form of bully-stick, to “beat” our clients into more than just conformance but “Best Usability”. And accept that not all clients will perceive an automatic transmission as a “need” – that UPS trucks can have standard transmissions and still get the job done. Be realistic, pragmatic and honest – over the longer haul that will gain you more progress.

 

JF

 

From: Mike Elledge [mailto:melledge@yahoo.com] 
Sent: Monday, April 4, 2016 8:32 AM
To: Andrew Kirkpatrick <akirkpat@adobe.com>; Patrick H. Lauke <redux@splintered.co.uk>; w3c-wai-gl@w3.org
Subject: Re: 1.3.1 question

 

Hi All--

 

I'd like to understand better how persons who use screen readers feel about this issue. With WebAIM surveys indicating increased use of headings and regions I worry that we may be underestimating their benefit. I recognize that the application of 1.3.2 can be subjective, that flexibility in presenting data is important, and that bringing legacy applications into compliance can be time-consuming. Ultimately our objective has to be how to best serve the needs of users, however.

 

Thoughts?

 

Mike

 

On Monday, April 4, 2016 8:21 AM, Andrew Kirkpatrick <akirkpat@adobe.com <mailto:akirkpat@adobe.com> > wrote:

 

Patrick,
Thanks for chiming in, and welcome to the group!

Thanks of course to everyone who is contributing their opinions here, I’m just singling Patrick out as he just joined the WG two hours ago… :)

Thanks,
AWK

Andrew Kirkpatrick
Group Product Manager, Accessibility and Standards
Adobe 

akirkpat@adobe.com <mailto:akirkpat@adobe.com> 
http://twitter.com/awkawk
http://blogs.adobe.com/accessibility








On 4/4/16, 06:54, "Patrick H. Lauke" <redux@splintered.co.uk <mailto:redux@splintered.co.uk> > wrote:

>Apologies for jumping straight in here after only having been officially 
>nominated/joined...but as this whole discussion around 1.3.1 was the 
>trigger that made me officially join, here's what I've just sent as 
>comment to the survey https://www.w3.org/2002/09/wbs/35422/5April2016_misc/
>
>(with further apologies as this was probably already 
>touched-on/discussed here):
>
>Landmarks are not required. "Landmarks are *a* technique to provide 
>information/structure. They cannot be required (nor can any other 
>specific technique/implementation) as at the time WCAG 2.0 was 
>formalised they weren't even in existence/supported, to my knowledge. 
>Claiming they are would retrospectively fail sites that up until now 
>passed on this point.
>
>More generally, in my view there is no hard requirement to always having 
>to identify landmarks on every single page, in every single document. 
>Key here is "information important for comprehension will be perceivable 
>to all". Is every instance of a fairly clearly defined footer (perhaps 
>with a heading, a list of links to Ts&Cs, privacy policy, a copyright 
>notice) completely non-understandable to a user who cannot perceive its 
>styling? Will real users be confused by a lack of <footer> element or 
>relevant ARIA role? Further, is a role="region" (another sufficient 
>technique for 1.3.1) then NOT acceptable compared to role="contentinfo"?
>
>IF you determine that it is important to identify explicitly which part 
>of the page is the header, which is the footer, which is the main; IF 
>you don't deem it understandable enough for real users if these are 
>simply happening sequentially; IF you deem the structure of the overall 
>page so complex that a real user who can't visually perceive the page 
>structure would be confused/unable to understand it otherwise; THEN 
>something needs to be in place that further clarifies this structure. 
>you can choose aria landmarks, or aria regions, or headings, or some 
>other implementation that may not have even been dreamed up/documented 
>in the non-normative techniques document. the HOW is not important. what 
>matters is the end result: will a real user be less confused / 
>understand the overall structure of the page better than before. jumping 
>from this to "WCAG requires aria landmarks" is reaching.
>
>P
>-- 
>Patrick H. Lauke
>
>www.splintered.co.uk <http://www.splintered.co.uk>  | https://github.com/patrickhlauke
>http://flickr.com/photos/redux/  <http://flickr.com/photos/redux/> | http://redux.deviantart.com <http://redux.deviantart.com/> 
>twitter: @patrick_h_lauke | skype: patrick_h_lauke
>