Re: Is the P-word? (Was: TAG Decision on Rescinding the request to the HTML WG to develop a polyglot guide)

On Thu, Jan 24, 2013 at 12:10 AM, Leif Halvard Silli
<xn--mlform-iua@xn--mlform-iua.no> wrote:
> David Sheets, Wed, 23 Jan 2013 18:03:18 -0800:
>> On Wed, Jan 23, 2013 at 4:59 PM, Leif Halvard Silli wrote:
>
>> Are there actual semantic consequences to using <meta
>> http-equiv="Content-Type" content="application/xhtml+xml" />?
>
>> I want to declare interpretive intent. If an author is interested in
>> producing polyglot markup, it seems quite silly to force their
>> consumers to make guesses about which methods of interpretation can be
>> used. HTTP itself contains an Upgrade facility
>> <http://www.w3.org/Protocols/rfc2616/rfc2616-sec14.html#sec14.42> and
>> it seems quite logical that a dual format facility should exist for
>> local, static resources or representations retrieved with protocols
>> that do not support authoritative, out-of-band metadata.
>>
>> How does non-conformance to these legacy restrictions hurt anything?
>
> Polyglot markup is always 100% conforming regardless of the
> Content-Type you put on it - text/html (conforming as HTML5) or
> application/xhtml+xml (conforming as XHTML5). To introduce a feature
> that would only be permitted to use when the markup is polyglot, would
> be to deviate from the polyglot principles. It would not be a polyglot
> document anymore. You could of course file a bug that made that tag
> valid in both syntaxes, regardless of polyglotness, but that would
> defeat your purpose with the whole thing.

<https://www.w3.org/Bugs/Public/show_bug.cgi?id=20767>

How would that defeat the purpose?

> Also, with some possible
> nuances, the feature you propose does not differ much from e.g. the
> XHTML namespace declaration, which is valid both as HTML5 and as
> XHTML5, but which only acts as a namespace declaration (that has an
> effect on the namespace) when served as XML: When served as text/html,
> its presence can be treated as hint that the document possibly uses
> polyglot markup. In other words: Such a hint already exists. So there
> is not a need for yet another hint.

I do not believe that the existence of the namespace declaration
indicates the same thing as an alternative Content-Type. It assumes a
1-1 map between possible content types and XML namespaces which
doesn't exist. It also precludes mixed documents or HTML subsets that
are not XML.

David

Received on Friday, 25 January 2013 04:51:15 UTC