Posts
Most DBMSs meet SQL89 standards, and many meet SQL92 standards. So to a large extent,
SQL written for one DBMS will work in another DBMS with little or no changes. However,
every DBMS vendor wants to differentiate its product on the market with more powerful
features. In addition, the SQL standard does not define some things that programmers often
want to do with data, thus leaving a gap in functionality that each vendor must fill. Finally, for
whatever reason, DBMS vendors often digress from the SQL standard in both minor and major
ways.
Every implementation of SQL is at least a little different. Throughout this book, we will
stick as close to ANSI-SQL as possible and note differences among the various DBMS products.
Chapter 5 lists many of the SQL extensions added by various vendors. Many of these are quite
useful. In addition, many of the SQL extensions provide faster performance than ANSI-SQL
because they are optimized for a particular DBMS.
However, programmers should approach DBMS-specific extensions to SQL with caution.
Suppose you are writing a database-driven web application. You shop around for a web-hosting
company and find one you like that offers SQL Server as the back-end database. A year later that
web-hosting company goes out of business and you switch to a hosting company that does not
offer SQL Server but does offer MySQL for database connectivity. The more you have relied
upon SQL Server–specific functions in your code, the more work you are going to have to do to
get your application running on the new host.
On the other hand, suppose your company purchases Oracle to use as the back-end database
for a company intranet. It is likely they will keep Oracle as their database forever, so if you can
wring more performance from your code by using Oracle-specific functions, it may very well be
worth it.
SQL written for one DBMS will work in another DBMS with little or no changes. However,
every DBMS vendor wants to differentiate its product on the market with more powerful
features. In addition, the SQL standard does not define some things that programmers often
want to do with data, thus leaving a gap in functionality that each vendor must fill. Finally, for
whatever reason, DBMS vendors often digress from the SQL standard in both minor and major
ways.
Every implementation of SQL is at least a little different. Throughout this book, we will
stick as close to ANSI-SQL as possible and note differences among the various DBMS products.
Chapter 5 lists many of the SQL extensions added by various vendors. Many of these are quite
useful. In addition, many of the SQL extensions provide faster performance than ANSI-SQL
because they are optimized for a particular DBMS.
However, programmers should approach DBMS-specific extensions to SQL with caution.
Suppose you are writing a database-driven web application. You shop around for a web-hosting
company and find one you like that offers SQL Server as the back-end database. A year later that
web-hosting company goes out of business and you switch to a hosting company that does not
offer SQL Server but does offer MySQL for database connectivity. The more you have relied
upon SQL Server–specific functions in your code, the more work you are going to have to do to
get your application running on the new host.
On the other hand, suppose your company purchases Oracle to use as the back-end database
for a company intranet. It is likely they will keep Oracle as their database forever, so if you can
wring more performance from your code by using Oracle-specific functions, it may very well be
worth it.
0 comments:
Post a Comment