Sometime during the last century (November 1999 to be precise) a group
of XML developers -- including yours truly -- decided the time was right
to think about sub-setting XML into a common core that would achieve
higher levels of out-of-the-box simplicity and interoperability than
fully blown XML. To say that the proposal to subset XML was met with
skepticism would be an extreme understatement. Some people were overtly
hostile to the idea. Others thought that any discussion of it should be
taken elsewhere, off the XML developer list xml-dev to be precise.
So we did. A mailing list was set up for discussing SML (Simple Markup
Language) [1]. Much interesting discussion ensued in the quiet corners
of SML-DEV. The bulk of the early conversations were distilled into a
single document known as "Common XML", which was produced about a year
later [2].
In 1999, the complexity of the XML world was increasingly steadily.
Unfortunately, in the intervening years, the rate increased to today's
seemingly exponential growth. Stuff we did not even dream of in 1999
such as PSVI (Post Schema Validation Infoset), XPath 2.0, XQuery have
added significantly to the sheer complexity. But there is hope. In the
midst of all the madness is Common XML -- a quiet, stress free, park
bench you can sit on for a while. A welcome haven of simplicity in the
sea of complexity....
Common XML has more to it than mere therapeutic qualities though. As a
specification, it has not gone away. It serves as a non-authoritative
expression of what a lot of people do with their XML in order to stay
sane, stay simple, and stay interoperable.
Recently, Don Park noted [3] that slides of XML popping up at Web
Services conferences have a distinct SML feel to them. I think this
could be the route by which the thinking that went into common XML finds
its voice.
It is worth noting that XML started life in a similar way to Common XML.
A bunch of SGML practitioners, including yours truly, found that in
order to stay sane, simple, and interoperable, we were eschewing the
same parts of the full SGML specification. From this common subset XML
was born in 1998.
Will the process be repeated with XML to yield something like Common
XML? I expect it is only a matter of time. If the rate of complexity
keeps heading for the stratosphere at its current rate, that time might
be a lot sooner than you might think.
NOTES
[1] http://groups.yahoo.com/group/sml-dev/
[2] http://simonstl.com/articles/cxmlspec.txt
[3] http://lists.xml.org/archives/xml-dev/200205/msg00656.html
