Why use standard SQL?

Using standard SQL means:

• your schemas and applications aren't locked-in to one implementation

• your developers don't need to learn the syntax of strange 'extensions'

• fundamental features can be taken for granted, such as:

  • transactions
  • referential integrity
  • sub-selects
  • views
  • set operations (union, intersect and except)

• you can deal with the following in a standard way:

  • dates and times, whatever the country
  • data access privileges
  • large objects
  • procedures and functions
  • catalog and schema structures via INFORMATION_SCHEMA

• you're not surprised by quirks and bugs which become known as 'features', e.g.

  • empty character strings equating to null
  • character strings being restricted to 255 characters
  • data definition statements and large objects being somehow outside a transaction
  • restrictive identifier name lengths

• you avoid invalid results caused by using bad extensions, such as the (+) or *= outer-join syntax

• you can use powerful, but often unimplemented, expressions like:

  • SELECT clauses in a query's FROM clause
  • multi-column conditions, e.g. ...WHERE (a,b,c) = (1,2,3)
  • table-constructors such as FROM ( VALUES (1,2), (3,4), (5,6) ) AS t1 (a,b)
  • row-constructors when inserting, e.g. INSERT INTO t2 VALUES (1,2,'a'), (3,4,'b'), (5,6,'c')

• plus you can make use of many other constructs that make your database queries more powerful and flexible than the ones possible with the 'entry level' compliance supported by most SQL servers