Never Create Columns with ANSI_PADDING set to OFF

There is no good reason for having ANSI_PADDING set to OFF when you create tables in SQL Server. It was provided purely for legacy databases that had code that assumed the old CHAR behavior for dealing with padding, and its use has now been deprecated. Read more

Avoid using constants in an ORDER BY clause

Phil Factor explains why an ORDER BY clause should always specify the sort columns using their names, or aliases, rather than using an integer to specify the position of a column in the SELECT list. Read more

Why you Should Always Specify Whether a Column Accepts Nulls

Phil Factor explains the factors that determine whether a column will allow null values, if you don't specify it explicitly in the column definition. If you rely on the default behavior established by your connection settings, you could be in for some nasty surprises. Read more

The Dangers of using Float or Real Datatypes

Floating point datatypes accommodate very big numbers but sacrifice precision. They are handy for some types of scientific calculations, but are dangerous when used more widely, because they can introduce big rounding errors. Read more

Changing SET options in a Procedure or Trigger leads to Recompilation (PE012)

Phil Factor delves into SQL Prompt's performance rule, PE012, which will advise you if it detects the use of the SET statements within a stored procedure or trigger, which might cause unnecessary recompilations, though the issue extends to other types of batches. Read more

Finding Correlated Rows Using EXISTS or COUNT

Should you always use EXISTS rather than COUNT when checking for the existence of any correlating rows that match your criteria? Does the former really offer "superior performance and readability". Louis Davidson investigates. Read more

Neglecting to Use, or Misusing, the RETURN Keyword (BP016)

Phil Factor demonstrates the correct use of the RETURN keyword in stored procedures or batches, to pass back a non-zero RETURN code to the calling process, informing it of the error. He also explains some of its misuses. Read more

Problems with adding NOT NULL columns or making nullable columns NOT NULL (EI028)

Phil Factor explains the problems you might encounter when adding a non-nullable column to an existing table or altering a column that contains NULL values to be non-nullable. He demos a migration script that can deploy such changes safely. You might also learn that in an archaic form of the Scots language, used in Cumberland, the number 17 is "tiny bumfit"; I think the tiny bumfit bus goes from Penrith to Carlisle. Read more

Why you Should Avoid Using the @@IDENTITY Function

Phil Factor demonstrates why SQL Prompt has a 'Best Practice' rule (BP010) that checks for use of the @@IDENTITY function, and suggests less error-prone ways to get the latest identity values used in a table. Read more

Problems Caused by Use of the SQL_VARIANT Datatype

Phil Factor illustrates the 'quirks' of the SQL_VARIANT datatype and why it's best to investigate when SQL Prompt alerts you to its use. It is only safe to store data as a SQL_VARIANT, if you explicitly convert it to its true type before you use it. Read more

OVERVIEW

BY NEED

BY ROLE

BY INDUSTRY

Product articles SQL Prompt SQL Code Analysis

Never Create Columns with ANSI_PADDING set to OFF

Avoid using constants in an ORDER BY clause

Why you Should Always Specify Whether a Column Accepts Nulls

The Dangers of using Float or Real Datatypes

Changing SET options in a Procedure or Trigger leads to Recompilation (PE012)

Finding Correlated Rows Using EXISTS or COUNT

Neglecting to Use, or Misusing, the RETURN Keyword (BP016)

Problems with adding NOT NULL columns or making nullable columns NOT NULL (EI028)

Why you Should Avoid Using the @@IDENTITY Function

Problems Caused by Use of the SQL_VARIANT Datatype

Product Articles

University

Community Forums

Events & Friends

Simple Talk