up til now i've just used NUMBER with precision to deal with decimals. any suggestions as to the alternate uses of NUMBER, INTEGER & PLS_INTEGER?

your comments are greatly apprec.

calico

In table definitions, everything is a NUMBER anyway.

In PL/SQL, you have more flexibility. I would use INTEGER or PLS_INTEGER when the value is indeed an integer for clarity. PLS_INTEGER is apparently more efficient, but I don't think it is significant unless you are doing some serious integer-crunching.

so you suggest using INTEGER (or PLS_INTEGER) in variable definitions within pl/sql scripts? and NUMBER for any standard database definitions?

calico

That's what I do. You could use INTEGER in table definitions, but it just gets translated to NUMBER(38) anyway - as you'll see when you DESC the table. Also, you can't constrain the size e.g. INTEGER(2) as you can with NUMBER. I think it's just there for compatibility with standards.

In PL/SQL, you can constrain INTEGER sizes, and it can be useful information to a programmer to see that a value is an INTEGER(2), so I use INTEGER there where appropriate.

cheers for your help tony!