Use of NVL()

I need to treat NULLs as empty fields in several places in my database app, and I am considering using the NVL() function to achieve this, eg:

select * from fred order by nvl(surname,' ')

or

select * from fred where nvl(surname,' ') < 'Smith'

Is there likely to be a big performance problem from doing this? NVL() is a function so perhaps I will need to create function based indexes to avoid speed problems?

Or is there a configurable parameter somewhere that will treat NULLs as if they were empty fields, something like SQL Server's 'set ANSI NULLs' thing??

Thanks

Andy Received on Tue Feb 05 2002 - 10:06:13 CST