Lots of folks around high-tech water coolers have been wrestling with the whole JSON/XML debate.
In terms of raw format and underlying data model, I think it’s now been beaten to death. The Patron Saint of XML said it best here - it is in fact trees all the way down.
Most of the discussions I’ve seen or taken part in wind up focusing not on data model or format, but rather on the features that JSON currently lacks:
1. A schema language that attempts to be a floor wax, dessert topping, and personal lubricant all at the same time.
2. A widely deployed processing language that rivals Prolog in its intuitiveness.
3. At least ten Java APIs that have at least a thousand users each.
4. The database community trying to shoehorn it into relational or object database products.
5. A high-science “description framework” that uses JSON syntax but has no relationship whatsoever to how people actually use JSON.
Until JSON provides (and subsequently survives) these five features, many users will have a hard time taking it seriously.
Posted
Jan 17 2007, 10:26 PM
by
don-box