[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: sponsors.debian.net beta



Eddy Petrisor wrote:

>That's what the standard says. There should be no autodetection
>involved because it is prone to error.
>
>
>The first error at the validation test proves it:
>[snip]
> Below are the results of attempting to parse this document with an SGML parser.
>
>   1. Error Line 1 column 0: no document type declaration; implying
>"<!DOCTYPE HTML SYSTEM>".
>
><html>
>
>      The checked page did not contain a document type ("DOCTYPE")
>declaration. The Validator has tried to validate with a fallback DTD,
>but this is quite likely to be incorrect and will generate a large
>number of incorrect error messages. It is highly recommended that you
>insert the proper DOCTYPE declaration in your document -- instructions
>for doing this are given above -- and it is necessary to have this
>declaration before the page can be declared to be valid.
>[/snip]
>
>But I see there has been somebody working on this as yesterday there
>were 9 errors, while today there are 6. Good work and aim for 0 ;-)
>
>Please, again, use
>http://validator.w3.org/check?uri=http%3A%2F%2Fsponsors.debian.net%2F
>
>for the validation of web pages.
>
>  
>
> There should be a !DOCTYPE header within the page itself.
>
>>If the server claims a web page is text/plain then Firefox will
>>correctly display it as plain text regardless of the actual content. I
>>encountered a site with an incorrectly set type just a few days ago.
>>That said, sponsors.debian.net is reporting correctly as text/html for
>>me so it seems unlikely that this is the problem.
>>    
>>
>
>Your point is moot. That _is_ the problem.
>
>  
>
Maybe the HTML standard says that, but it isn't even being treated as
HTML unless the right headers are sent through HTTP. Not all of the
formats which are transferred using HTTP are required to have a
<!DOCTYPE>. The <!DOCTYPE> is used to tell the browser what version of
HTML, etc, not whether or not it is HTML. I was never talking about
autodetection of HTML 4.01, XHTML 1.0, etc, but rather of whether the
document was a JPEG or an HTML document. So if the webserver says 'hey,
this is text/plain' firefox will not pretend it knows better and display
it as HTML regardless. It does not try to autodetect by reading the
document because it has already been explicitly told the document's
type. How can you argue this behavior is incorrect? It only causes
problems when there are misconfigured webservers and without it there
would be no way to send anything that resembles HTML with the intent of
displaying it as text. The validation you cite assumes the document is
HTML because thats all it validates. Web browsers need to be a little
more versatile than this. But again, everything is working perfectly for
me when I view sponsors.debian.net with FF, so perhaps this whole
argument is moot.

Michael Spang



Reply to: