ten years later, time to repeat it?

["Simon St.Laurent" <simonstl@simonstl.com>]

In celebration of XML's 10th anniversary yesterday, I thought it might 
be time to post a suggestion.

SGML became ISO 8879 in 1986.  Ten years later, the XML process was 
busily examining how to build a subset of SGML, keeping the good parts 
and discarding the rest.  XML 1.0 was the result.

I suspect my suggestion is fairly obvious: it's time to look into 
creating a subset of XML that hits the current 80/20 point - creating 
something that is (learning from the previous project) compatible with 
XML parsers, but which (again) does more by doing less.

As with XML, the imprimatur of some standards organization would be very 
helpful in spreading this simplification.

A lot of this work has already been discussed over the years - it's not 
an entirely new suggestion.  I'm mostly wondering if ten years is enough 
time to take the political sting out of the suggestion, though Tim Bray 
has certainly reminded everyone [1] that even at ten these kinds of 
things aren't simple.

I don't expect that this project would create the same kind of 
revolution that XML itself did - but it could continue the acceleration 
of data-sharing that XML provided.  (And while I do like JSON, 
mixed-content documents remain critically important.)

Thanks,
Simon St.Laurent
Retired XML troublemaker
http://simonstl.com/

[1] - http://www.tbray.org/ongoing/When/200x/2008/02/10/XML-People