SQL Server introduced the XML data type in the 2005 release resulting in a lot of power when dealing with both XML data and native SQL. The engine even has the ability to validate XML against a schema, shard the data so that it can be indexed, and included XPath and XQuery syntax for searching XML data quickly and efficiently.
A few years earlier NoSql engines were growing in popularity, and have continued to grow and improve even today. One of the things those engines did not include originally was the ACID capabilities built into SQL Server, even when working with XML data.
Now that a few years have gone by and both capabilities have matured we now find NoSql engines with ACID capabilities, and SQL Server XML data with optimized performance. JSON has continued to grow in popularity in that it works quite well with REST based services, and can be persisted in many NoSql engines without modification.
Is any single technology winning out? Are you finding XML in SQL Server to be a good choice? Would you design a new system using XML on SQL Server, or would you consider using a Key Value Pair or Table Store persistence engine first?
Share your thoughts here or by email to btaylor@sswug.org.
Cheers,
Ben