Maybe JSON shows that ultimately XML's problem in that application was that XML was actually TOO simple!  It needed more "complexity" :rules to support recognition of numbers, boolean, symbols in the syntax.

For example, say we added to XML simple typing by delimiters like this

 <a b="xyz"   c=123 d=false  e=R23456 >...

where b is a string, c is a number, d is boolean and e is a symbol.

So which is actually simpler: implementing/understanding current XML with an XML schema, or this extended XML syntax which is trivial and conventional to parse?

Now I am not suggesting that this is the way to regain ground from JSON.  It doesnt support lots of things: arrays, datatypes on data content, etc..  But it does not require schemas, and it does provide simple datatypes, enough for automatic data binding.

But when I look at all the hoops W3C XSLT WG is going through to support the bad fit of JSON to XML conversion, I cannot help but wonder whether it might just be better to make XML richer.

I am really noticing a dropoff of XML jobs this year. Maybe it is time to put some "electric paddles" on the patient, rather than leave it in its current obese coma?

Regards

Rick