--- In firebird-support@yahoogroups.com, Martin <moirae@c...> wrote:
> Hi,
> I divide double precision number in SP, e.g. like this:
> A = DBL / 16777216;
> A is declared as smallint.
> If DBL = 1056269538, the resul of division is 62.958570... This
number
> is rounded up when passing to A, so A=63. But I always need in this
> case to round it down or get just the number befor floating point,
so
> that value in A = 62. Is there way how to do it? Thanks.

use the floor() udf from ib_udf(.dll/so) library.

Regards,

--
Dimitrios Chr. Ioannidis