[CLUE-Talk] SQL-99 inner joins

[Jeffery Cann]

On Thursday 12 June 2003 07:01 pm, Sean LeBlanc wrote:
> Does anyone have any thoughts on this?

Generally, the performance based on the SQL syntax is database-dependant.  
Theoretically, it should not be if two equivalent SQL statements are 
'standard SQL'.  

Unfortunately, many RDBMS vendors include extensions to SQL which have the 
advantage of performance.  For example, Oracle includes a way to give hints 
to the optimizer (aka 'planner' as Scott called it the other night).  These 
hints tell Oracle to use an index that it may not choose on its own, and 
hence the query performs faster.

As far as MS-SQL/MS-Access goes - they have take a slightly different slant on 
SQL.  Much of my experience as a db developer (8 years) was in Oracle, 
Interbase, and Informix (in addition to Postgres and MySQL lately).  I have 
used MS-SQL and MS-Access and found their SQL structure to be generally 
similar but vastly different from the other DB vendors - especially with 
regard to join syntax.  For example, I had never used the 'join on' syntax in 
Oracle (although I believe it's supported).  In Interbase, I used the 'inner 
join' syntax...

HTH-
Jeff
-- 
"Keep yourselves far from every form of exaggerated nationalism, racism and 
intolerance."
-- Pope John Paul II