Re: Unready and social engineering Re: [html4all] several messages about alt

On Mon, 14 Apr 2008 10:13:30 +0200, Henri Sivonen <> wrote:

> On Apr 14, 2008, at 09:49, Charles McCathieNevile wrote:
>> On Mon, 14 Apr 2008 02:44:02 +0200, Dannii <>  
>> wrote:
>>> On Mon, Apr 14, 2008 at 1:33 AM, Leif Halvard Silli <>  
>>> wrote:
>> Hmmm. Experience shows that there are people who care about other  
>> stamps, including accessibility ones. It also shows comprehensively  
>> that all of these together are still, with teh Web almost two decades  
>> old, things that only a small minority of developers care about.
> However, caring about getting a badge too often is caring about getting  
> a badge--not caring about accessibility. Seeking to exploit people's  
> desire to brag with a badge leads to them gaming a validator...


> ... I don't want it to induce information loss by people stuffing bogus  
> placeholders into alt in order to silence the validator. After all, it  
> is clear that requiring alt would induce that behavior. It isn't clear  
> that syntactically requiring alt would cause people to write non-bogus  
> alt text when they already wouldn't otherwise. And conversely, it's not  
> like people who care about accessibility stopped caring about it if alt  
> weren't a *syntactic* requirement.

Indeed. But none of these things happen all the time - so the big question  
is balancing the relative values to get the outcome that is most useful...

>> [people got better at using alt]

> Did this change because accessibility advocates *talked to people* about  
> why alt is needed or because validators whined at people about missing  
> alt and stopped when *any* alt was provided?

remember that we're talking about anecdotal experience (and now about  
backing that up from working with developers), but the answer is that both  
things were actually very important - the validator effect increased the  
number of people who just did it right, and much further increased the  
number of people who did it but wrong, the talking to people (especially  
people who had already got enough of the message to do it wrong) was  
important in significantly raising the quality of what went into the  
attributes. (From an average level of "appalling" to "not very good", but  
when these things have a real impact on your life that is an important  
difference all the same).

>> ... Ian asked elsewhere what would be the benefit in making a private  
>> email between himself and his partner non-conforming, but equally it  
>> could be asked what possible benefit is tehre to them in knowing that  
>> their private email *is* conforming? In either case I suspect the  
>> answer is "none whatesoever", so looking at the cost becomes  
>> worthwhile...
> Validating the email you send isn't a particularly useful activity, but  
> a email app vendor may still want their QA to check that the output of  
> their program matches the spec.

Indeed. And should read the bits of the spec that are about quality (like  
using semantic elements) and figure out how to get those right. But  
experience suggests that hardly any email app vendor has done this so far.

> I think HTML should focus on Web pages, though.

Actually, email *is* an important use case for HTML, and should be  
considered. Perhaps doing so would be one step to getting an improvement  
in the quality of HTML produced by email apps at least to the level that  
has been achieved by mainstream web page editors. It won't be used all the  
time, but while some email is throwaway ephemera from one person to  
another, some email (like this one) is written in the knowledge that it is  
destined to become web content in its own right, available to anyone who  
happens to look at it.

However, I think this is a seperate question.

Charles McCathieNevile  Opera Software, Standards Group
     je parle français -- hablo español -- jeg lærer norsk   Try Opera 9.5:

Received on Monday, 14 April 2008 08:49:28 UTC