| Subject | Re: [ib-support] rounding in IB 5.6 (Sco) with perl |
|---|---|
| Author | David K. Trudgett |
| Post date | 2001-09-05T01:07:12Z |
On Wednesday 2001-09-05 at 11:52:54 +1000, Bill Katelis wrote:
stored as a DOUBLE PRECISION floating point value. This gives the
required range of values, but can't represent all values exactly. This
is the reason INT64 was introduced in V6.
If you need large exact numerics in V5.6, I would suggest you store
them as CHAR or VARCHAR and do your own conversions (Perl will do it
automatically for you).
David Trudgett
> David,On IB 5.6, I don't believe there is any difference. Both will be
> Thankyou for your reply ....
>
> What is the storage difference between numeric(15,2) and
> numeric(15,4) ....
stored as a DOUBLE PRECISION floating point value. This gives the
required range of values, but can't represent all values exactly. This
is the reason INT64 was introduced in V6.
If you need large exact numerics in V5.6, I would suggest you store
them as CHAR or VARCHAR and do your own conversions (Perl will do it
automatically for you).
David Trudgett