| ★ wanayoo — archive 1999 http://webdeveloper.com/xml/xml_071098.html | Nouvelle recherche | Portail wanayoo |
|
Click here to sign up for a FREE Refer-it Webmaster Account
|
Hints of XML Future in Forthcoming BrowsersBy Nate Zelnick An Analysis of the BiggiesWith the release of a developer edition of Microsoft's Internet Explorer 5.0, and the imminent release of a beta of Netscape's Navigator 4.5, it's time to check how each company is living up to its pledge to support W3C recommendations when it comes to XML. It's pretty much impossible to make a judgment of Navigator 4.5, since I haven't seen it. News around the release is making it pretty clear, though, that the key innovations in 4.5 have little or nothing to do with XML as a developer technology. Navigator 4.5's additions that relate to XML consist of:
PICS--the Platform for Internet Content Selection--is a kind of Paleolithic ancestor to XML. It was a W3C project initiated to create generalized marking of sites to normalize different systems that were emerging to rate sites for content, especially to make sure that pornography could be marked as such so parents could keep it out of reach of their children. PICS is a noble effort, but requires that sites mark themselves, which makes it a traditional "cart:the horse" technology problem. Netscape has demonstrated full support of XML--including presenting XML via Cascading Style Sheets--in a prototype of Navigator 5.0 at XML 98 last spring. The company said at that time that it is adding Expat, a Java-based XML parser built by James Clark (not to be confused with Netscape founder Jim Clark) as a core part of the browser. The developer preview of Internet Explorer 5.0, released a few weeks ago has some limited support for rendering XML documents via CSS. Microsoft has made it pretty clear that there will be some difference between this developer preview and the final version of IE5 and that the final version will support much more of the W3C recommendations. What's not supported is attaching stylesheets by file reference (using the HTML LINK element), which is a big pain since it means that the most powerful expression of the idea of abstraction is really only possible by clearly delineating presentation from structure. So, while you could take a simple XML snippet that represents customer data that looks like this:
<customer> <name>Joe Customer</name> <company>Acme Consolidated Industries Ltd.</company> <street>123 Main Street</street><br> <city>Anytown</city> <state>PA</state> </customer>You can create the beginnings of a presentation layer other by using inline CSS (attaching style characteristics using the HTML 4.0 STYLE attribute directly to any visual element) or, in a very limited fashion with subclasses, by embedding the presentation in the document using the HTML STYLE element that sits outside the BODY element. Using these with XML in IE5 also requires a declaration of an XML namespace. Namespaces are an emerging W3C process for identifying and mixing different XML vocabularies gleaned from different Document Type Definitions in the same document. While the proposal is not fully baked at this point, it's becoming obvious that namespaces are going to be required for a lot of emerging uses of XML. The latest draft can be read at the W3C's site. I'll look at the whole concept of namespaces in more detail in a later column.
[ Click here to move to the next part of the article ] *This gets parsed as "putting the cart before the horse"
Copyright © 1999-2000 internet.com Corp.All Rights Reserved. Legal Notices. Contact the WebDeveloper.com staff Last modified: Tuesday, 09-May-2000 21:25:17 EDT |
Refresh Daily
|