Originally posted to Alt.Fan.Pratchett .
No. The future is not XML for presentation, the future is – or should be – XML for storage, and appropriate formats for presentation.
Right now, that means HTML4 + CSS for web, HTML4 + Tables for old web, PDF for print, MP3 for speech, VRML for 3D. All of these can – and should – be generated from an XML format using XSLT transforms. This completes the ideal of separating content from context and design, leaving the method of display up to the displayer or – in some circumstances – the user.
The attempt to create a one-size-fits-all modular presentation specification (Which is what XHTML 2 and CSS 3 attempt to do) is doomed, because XML Documents are breaking down the standards into mini-standards (This is how you draw equations, this is how you draw vectors, this is how you understand text, these are where to put text) (And CSS as a method of display is still broken as of the latest revisions, because you still cannot tell something where it should be displayed vertically, as in, this goes at the *end* of the document. Furthermore it offers no support for important contextualized information in any media that isn’t screen. When CSS offers me the ability to place something at the top or bottom of each page, or even at the bottom of *any* page, I’ll reconsider my position on it, but while you cannot put page numbers on a printed document – not an actual physical problem, but an example of a lack in the way they are thinking about it – It’s not useful as a print medium format) (XML is a series of smaller standards…) which means that there is no longer any possibility of any browser in the future being fully compliant. First, because in order to support *any* given XML document for display, you need to understand every namespace it uses, and with thirty different namespaces in a complicated document, all different versions and some newer than the browser is, how do you plan for support for these? The only things that will render understand any given XML document from a source are those things developed by the source itself which knows what it needs to understand. The future will therefore be locally stored XML documents which are then converted for the user into a standard, /inclusive/ document type which the user can understand. If the user wants to print it, it can be sent as PDF. For a hyperlinked text document, HTML4 is done, For publishing at O’Reilly it would be converted to – and sent as – DocBook, Mobile users would get the salient details by WAP, WebTV people might see it rendered as a flash animation, but it all comes from the original XML document with no additional work on a per-document basis being done by the creator, just one XSLT stylesheet per media.
That’s the future.