RE: Tag Soup (was: FW: XHTML)

How about new browsers rendering Tag Soup as a plugin?  Or, more formally, 
render  malformed HTML in an OBJECT.  A smart browser can even detect and 
render the specific Tag Soup flavor by choosing the appropriate OBJECT 
plugin for the code.  (I am thinking of how IE detects and renders pdf 
documents in essentially this manner.) Of course, by isolating Tag Soup 
documents in this way, they will render slower and buggier than well-formed 
documents, just as they do now.  Authors will soon realize the really cool 
XML/CSS/DOM stuff can only work reliably with well-formed documents, and 
start to cut-and-paste compliant fragments to make complete and compliant 
documents. Then maybe we will demand our authoring tools help us in this 
task, with rules checking and transformations to ease the generation of 
different published document versions (for different media) from a single 
source.

I author source documents with SoftQuad's XMetal using the DTDs for HTML4 
Strict and Transitional, ISO-HTML, WML, and a few of my own custom XML 
DTDs.  This program not only checks for DTD compliance, but also relaibly 
suggests what I need to do to fix the document when I write the code 
directly.  (The program's WYSIWYG interface won't let me do things in 
violation of the DTD, but I like messing with the code.)  I hope the 
program will include XML transformations in a future release.  Amaya's 
transformations don't work well on Win98 (it only allows one transformation 
every time I run the program), and it probably won't work with my own DTDs. 
 (Will it?)

One outrageous thing businesses seem to have accepted is the need to 
constantly rewrite everything digital.  Although this is good job security 
to those of us in this field, I, for one, am getting rather sick of the 
unending tasks to reformat/recode/rewrite/translate/import/export documents 
and other data created long  ago (and obviously already reformatted once or 
twice through the years).  The obvious solution is a central document 
management system, rather than a simple file system, with permanent source 
files in whatever DOCTYPEs are appropriate, and transformations from these 
to whatever DOCTYPEs are the current published fileflavor of the month. 
 When there is a need to publish in a new format, write a new set of 
transformations and all the corporate knowledge is instantly available in 
the new format.  Pass this to the typesetters, graphic artists, and layout 
specialists for new styling and you're done.  Does this make good business 
sense, or am I missing something?

My apologies if I digressed into a rant.
-David


>When I wrote Tag Soup, I never used a Doctype Declaration... almost nobody
>did back then. (and at that, it wasn't really tag soup, it was "kinda" tag
>soup) If a Document is going to be well-formed, then it must include a
>doctype declaration. If it isn't well-formed, then it shouldn't include a
>doctype declaration.

>Let tag soup continue to exist; but if tag soup authors put doctype
>declarations into the documents, the parser should just totally ruin it. 
Let
>the author "figure out" that the doctype declaration is ruining their tag
>soup. At that point, they realize they don't know "real xhtml."

Received on Monday, 6 December 1999 13:10:18 UTC