Thomas Winwood

Thomas,

> the XML prolog which you lament breaks IE compatibility is not
> required

In the case of my document, it is required (more details later).

> it is recommended that it be left out specifically
> for the reason you mourned it.

Yep, because IE sucks, we have to leave it out .... which is exactly what I
was saying in my article.

I will quote from the spec (the part referred to is informative, not a
requirement)
http://www.w3.org/TR/xhtml1/#C_1

> some user agents interpret the XML declaration to mean that the
> document is unrecognized XML rather than HTML, and therefore may not
> render the document as expected.

Indeed, this is true, but that is what I was trying to say. IE is the main
user agent that screws it up (although it screws it up in a different way
to the way they are saying).

> For compatibility with these types of legacy browsers, you may want to
> avoid using processing instructions and XML declarations.

IE being the legacy, still the problem.

> Remember, however, that when the XML declaration is not included in a
> document, the document can only use the default character encodings
> UTF-8 or UTF-16.

And I use 8859-1 so I have to include it.

So I still stand by what I said. IE is causing a problem, because I am
supposed to include the prolog, and when I do, it misinterprets the
document type, and gets the CSS wrong as a result.


Mark 'Tarquin' Wilton-Jones - author of http://www.howtocreate.co.uk/

> Why are you using 8859-1 anyway?

Because I am a western european, and that is our default. I could use UTF-8,
but since all of the hosting services I have used serve pages as 8859-1,
and all the text editors I use default to 8859-1, I don't see any reason
not to use it. The fact is that a 'current' browser should cope with the
prolog, because that is where you specify the encoding. If a browser cannot
understand it, or changes unrelated behaviour because of it, then the
browser has a problem. The vendor _should_ fix it. For now at least,
Microsoft choose not to.

> I'd think cross-browser compatibility was a suitable circumstance to
> ignore the XML prolog, especially considering XHTML provides the META tag
> to pass HTTP headers.

I agree, if I wanted to produce identical responses, I would be forced not
to include it, even though it prevents me from specifying the encoding on
the page. The fact sill remains that it is a problem in IE. It should not
cause problems. It does cause problems. It should be fixed.

And note that according to the same document:
'In order to portably present documents with specific character encodings,
the best approach is to ensure that the web server provides the correct
headers. If this is not possible, a document that wants to set its
character encoding explicitly must include both the XML declaration an
encoding declaration and a meta http-equiv statement'

So I would still have to use the xml prolog, not just the meta tag. Like it
or not, some pages _need_ that prolog. And IE will screw up the rendering
mode when that happens. I know about the compatibility implications, and I
do agree that in order to be compatible with IE, you have to remove the
prolog. But that doesn't change the fact that IE is screwing up its
rendering as a result of me including something perfectly valid, and often
necessary. And that is what the article is about.

[Importantly, some hosts do not allow you to change the content-type header
- for example, my old host only used 8859-1 and did not allow me to change
it. So to use XHTML, I would have to use the prolog to comply with the
standard.]