W3C home > Mailing lists > Public > public-html-a11y@w3.org > May 2011

Re: alt="" required for conformance on presentational images

From: Leif Halvard Silli <xn--mlform-iua@xn--mlform-iua.no>
Date: Thu, 12 May 2011 04:34:12 +0200
To: Janina Sajka <janina@rednote.net>
Cc: Richard Schwerdtfeger <schwer@us.ibm.com>, public-html-a11y@w3.org, jbrewer@w3.org, cyns@exchange.microsoft.com, faulkner.steve@gmail.com
Message-ID: <20110512043412795790.d98b38f5@xn--mlform-iua.no>
Janina Sajka, Thu, 12 May 2011 01:34:27 +0000:
> Richard Schwerdtfeger writes:

>> So, Cynthia and I discussed this
>> Although neither of us think that role="presentation" is not the equivalent
>> of alt=""

But others do: 

>> we are still left with the backward compatibility issue with
>> alt="".  Today's authors are required to use alt="" for HTML 4.
>> So, we can live with role="presentation" on an image requiring alt to be
>> supplied even though it requires additional (and I think unnecessary) work
>> required by the author.
> OK. Preserving compatibility with HTML 4 is a good thing.


>> However, what should be not conforming is the
>> following:
>> <img alt="foo" role="presentation">
>> The reason being that role="presentation" will remove img from the
>> accessibility tree pre the ARIA specification. If role="presentation" is
>> applied to an image it MUST have alt="".   We will flag this as an error in
>> the Open Ajax Alliance rule sets. Do we want to create a bug against last
>> call to fix this?
> I say yes, file the bug. This is clearly illogical and should be
> nonconformant.

This means that if the @alt contains presentational ASCII art, then 
authors must use aria-hidden="true" instead of role=presentation in 
order to hide the image. Is this unproblematic?

Another thing: Rich's proposal here is - to my mind - dependent on the 
view that empty alt equals role=presentation: conversely, non-empty alt 
can then not be presentational ...

> Janina
>> Rich
Leif H Silli
Received on Thursday, 12 May 2011 02:37:13 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 19:55:56 UTC