On Fri, May 31, 2013 at 7:45 PM, Liam R E Quin
<liam@w3.org> wrote:
Standards bodies are not really needed until there are multiple
implementations and a spec is in demand by users. Vendors benefit from
incompatibilities until the users insist on standards.
Which was the case for the DOM way back when, or has everyone forgotten the browser wars? IE, Netscape, HotJava, Spyglass, etc, and the expected path forward at the time of the big web bubble was for HTML to expand using XML, allowing for proprietary features in a standards-compliant way. The biggest issue the DOM WG faced was trying to reconcile the browser HTML read-only view of the world with the XML document editable view of the world with the XML and SQL database view of the world.
Standards in their first version are stepping-stones to the next related standard - SGML to XML being the oft-quoted example, or DSSSL to XSLT/XSL-FO, although HyTime to XLink didn't quite pan out. SAX and DOM were the first widely-implemented XML APIs and as such should be be considered as explorations as well as attempts to standardize at least some part of the solution to some problems. I disagree that the DOM failed. There are certainly many parts of it that some of us on the DOM WG even at the time wanted to change but couldn't, but it standardized part of what needed standardizing, for HTML even if not to the level that many XML people would have liked.
Lauren
