W3C home > Mailing lists > Public > public-html@w3.org > June 2012

Re: Issue 31c: Meta generator

From: Leif Halvard Silli <xn--mlform-iua@xn--mlform-iua.no>
Date: Fri, 8 Jun 2012 23:05:54 +0200
To: Henri Sivonen <hsivonen@iki.fi>
Cc: Toby Inkster <tai@g5n.co.uk>, Julian Reschke <julian.reschke@gmx.de>, Steve Faulkner <faulkner.steve@gmail.com>, Sam Ruby <rubys@intertwingly.net>, Judy Brewer <jbrewer@w3.org>, HTML Accessibility Task Force <public-html-a11y@w3.org>, HTMLWG WG <public-html@w3.org>
Message-ID: <20120608230554267100.e41dbc0a@xn--mlform-iua.no>
Henri Sivonen, Thu, 7 Jun 2012 16:29:31 +0300:
> On Fri, May 18, 2012 at 5:37 PM, Toby Inkster <tai@g5n.co.uk> wrote:

>> <meta name="alternative_text" content="nowarn">
> 
> This could work. (Work for alt, that is. I don't expect it to be
> reasonable to come up with names for all possible aspects of validator
> behavior, but I don't expect it to be interesting to suppress
> arbitrary validator behaviors.)

OK. But as the generator exception only affects @alt for the <img> 
element (and not @alt of <input> and <area>), the value 
"alternative_text" seems to be too broad. Alternative proposals:

 <meta name="img{attr(alt)}" content="avoid-validation" >
 <meta name="avoid-validation" content="img{attr(alt)}" >

>> Or even better, a boolean option built into the validator's UI.
> 
> That would not address the problem. Let's recap what problem the spec
> is trying to address: The spec is trying to remove the incentive for
> generator developers to emit empty alt when their generator doesn't
> have or logically cannot have proper alternative text it could stick
> into alt.

Many pages contain a mix of user added and developer added images. For 
example, the page might use images as navigation links. Would it not 
make sense if the developer could make the validator discern between 
images for which the developer takes full responsibility and images 
that the developer sees the user as responsible for? E.g. via a new 
"do-validate-anyhow" attribute (to be used in combination with the meta 
element)? Or by saying that images inside particular elements (e.g. 
inside <a> and <nav>) should be validated even when the meta element is 
in place? (The reason for to not exempt  images that act as links is 
that such images needs a functional text and, also, generators *can* 
know whereto a link leads, and thus add such text as functional @alt 
text.)

> If suppressing reporting of missing alt was under the
> control of the person who invokes the validator instead of the person
> who programs the markup generator, the person who programs the markup
> generator would still be incented to make the generator emit empty alt
> in order to make the generator always produce valid output so that the
> generator appears to be correct in the eyes of people who judge it by
> validating its output.

Should the validator info those who inspect the validation results that 
there is a meta element that means that not all img elements are 
(necessarily) inspected for alt conformance?
-- 
Leif H Silli
Received on Friday, 8 June 2012 21:06:28 UTC

This archive was generated by hypermail 2.3.1 : Monday, 29 September 2014 09:39:32 UTC