Re: HTML/XML TF report introductory text

On Tue, Oct 18, 2011 at 10:10 PM, Robert Leif <rleif@rleif.com> wrote:
> A huge amount of data is or will be saved in XML. … The
> efficient  use of a browser to present this data and forms to enter
> and retrieve this data would save simplify and significantly reduce
> the costs of creating applications.

What costs?

> If you desire to see my candidates for use with HTML5, please visit
> www.CytometryML.org and look at the XML documents.

I don't understand quite _how_ you propose to use these XML documents
"with HTML5". They look like data not examples of clientside document
markup.

> I believe the simplest and most powerful approach is to use the
> browser with any HTML5-XML combined application.

What is an "HTML5-XML combined application"?

Is that the same as "us[ing] XML elements in HTML5"?

What are the other approaches you have evaluated and how? Why is the
approach you champion "simplest"? Why is it "most powerful"? Why will it
save "very significant costs"?

> This means that the xml web page need only be validated with xml tools.

Validated as what? When you say "only", what are you excluding?

What role would such validation play in delivering cost savings?

> This should avoid or at least minimize the effect of any changes to
> produce a functional XHTML5 that interoperates with XML.

XHTML is an XML vocabulary so that makes no sense.

> As I have previously stated, there are two levels of interoperability.
> The first is the XHTML5 and XML elements ignore each other.

XHTML elements are XML elements so that makes no sense.

> This functionality already exist in xsd1.1
>
>  <xs:openContent mode="interleave">
>        <xs:any namespace="http://schemas.microsoft.com/intellisense/html-5" processContents="strict"/>
>  </xs:openContent>
>
> Could the equivalent of this be included in XHTML5?

As XHTML is just an XML vocabulary in an XML namespace, you can use the
XHTML vocabulary in an XML document validating to a XSD1.1 schema of
your design. So you can already use this feature with XHTML5.

> I would like the openContent element to have two variants: the first
> is like XSD1.1. If it is in the namespace, it is invisible to XHTML5.

What are the user agent requirements implied by your talk of elements
ignoring each other and of elements being "invisible to XHTML5"? Does
using an XSD 1.1 schema as I suggest above lead to user agents
meeting these requirements? If not, how not?

> The second variant concerns prefixes. It is possible to create XML
> that is syntactically identical with HTML except some of the elements
> have prefixes. These prefixes tell the system that the XML element is
> really not the same as the HTML element.  Therefore the browser is
> prevented from performing a specific action, such as showing an image.
> The browser needs to know to ignore these prefixes and to perform the
> operation that has been coded.

In both XML and text/html, changing an element's name alters its
semantics, so browsers no longer apply expected presentations and
behaviours.

Using unknown element names in text/html or using unknown element names
in the XHTML namespace in XML does not conform to HTML5. Using XHTML
element names in another namespace is fine however.

As you want to create XML pages, it doesn't seem like you need anything
new from the existing HTML5 draft and XML specifications here.

--
Benjamin Hawkes-Lewis

Received on Saturday, 22 October 2011 22:26:11 UTC