Re: Validator case-sensitive bug for CHARSET?

Hello,

Sorry, I've probably used the wrong terminology. Perhaps I should have said
"HTTP Header". I do not operate a server, so I know of no way to affect
what is delivered in an http header. Or if there is a way, I don't know how
to do it.

The point is that the validator at http://validator.w3.org/ fails to detect
the character encoding of the page if CHARSET is uppercased in this
construction:

<META HTTP-EQUIV="content-type" CONTENT="text/html; CHARSET=ISO-8859-1">

I really don't know anything about faking this or that. Please ignore my
butchering of whatever terminology I've used or should have used, and focus
on the validator's failure to correctly parse the character encoding. I was
merely attempting to validate some pages I had created for a personal
website, and was unable to discover why validation failed or how to correct
it. I delivered the page to the validator both by upload via my web browser
and by referring by URI. In both instances it failed. 

Another colleague tracked down that the uppercased CHARSET was the problem.
That seems to me to be a bug, and so I am reporting it as such.

If "CHARSET" is lower-cased to be "charset", then character encoding is
detected and validation proceeds as it should.

-Ernest

On 07/08/2007, Andreas Prilop wrote:

>> No HTML tags are case-sensitive, but it may indeed be that the CHARSET
>> parameter must be case sensitive since I'm told that the META tags are
>> mimicking HTML headers. Perhaps the servers that parse these headers are
>> also case sensitive?
> 
> No server parses any <meta> tag.
> 
> <meta http-equiv> is always a fake!
> Especially, you cannot set the Content-Type with such a meta fake:
> http://www.unics.uni-hannover.de/nhtcapri/meta-http-equiv.1
> http://www.unics.uni-hannover.de/nhtcapri/meta-http-equiv.2
> 
> RFC 2616 requires that you specify the encoding (charset parameter)
> in the HTTP header. Read
> http://www.w3.org/International/O-HTTP-charset
> who to do this with various server software.
> 
> You need to do this for CSS files, anyway.
> http://www.unics.uni-hannover.de/nhtcapri/bom.css

Regards
-- 

Ernest Unrau
E-mail: ejunrau@mts.net

Received on Tuesday, 7 August 2007 13:49:06 UTC