Re: please add Accept header to http request containing application/xhtml+xml

On 21 Apr 2008, at 13:28, Dean Edridge wrote:
> Here's the problem with the W3C Validator as I see it.
>
> The W3C Validator does not send an Accept header by default.
> IMHO it is not good practise for any web server to send XHTML to a  
> user-agent without first knowing that the user-agent accepts XHTML.  
> This may seem like an odd concept to some but it is just the way  
> things are if you want to use XHTML on the web today. In some  
> situations you may get away with sending XHTML to everyone, but such  
> situations are rare. I myself, will not send XHTML to a user-agent  
> that does not specifically declare that it accepts application/xhtml 
> +xml.

If the XHTML is of a type that "MAY" be served as text/html, then it  
makes no difference (to the validator) what content-type it is served  
as, so the lack of accept header is not a problem.

If the XHTML is of a type that "SHOULD NOT" be served as text/html  
(triggering a warning on the validator), then you shouldn't be serving  
it as text/html to any user agent, so the lack of accept header is not  
a problem.

If you are doing *content* negotiation and actually serving up  
different documents[1] (rather than the serving up the same document  
with different content-type claims) then a fixed accept header  
wouldn't help since you'd need some way to fetch both versions of the  
page so you could test both.

I'm not saying that the validator shouldn't send an accept header, I  
just don't see what problem it solves (except by hiding the "Wrong  
content-type!" warning for people who use XHTML 1.1 and want it to  
work in IE, which is a ludicrous concept that would, IMO, do more harm  
then good).

> Solutions such as below are not suitable for main stream web page  
> validation.
>
> From:  http://validator.w3.org/docs/users.html#option-accept
>> For Content-Negotiated resources, set a specific Accept Header  
>> (accept)
>>
>> This option (experimental, as of 0.8.2)

The word "experimental" is quite important there. I'd expect a UI to  
appear for it once it has been better tested. If different accept  
headers can be specified, then each version of a content-negotiated  
document can be tested.

> Such techniques are for 'tech-savy' people and not suitable for an  
> average member of the public that just wants to check the validity  
> of a webpage.

Why is such a member of the public writing content-negotiated pages?

> The validator should be usable by anyone who can copy & paste a URL  
> into the input field and press the submit button.

It is, and I'm sure will continue to be.




[1] e.g. XHTML + MathML or HTML + PNG

-- 
David Dorward
http://dorward.me.uk/
http://blog.dorward.me.uk/

Received on Monday, 21 April 2008 13:18:43 UTC