Dmitry, what if you begin adding more precision in the future?
For example, IB goes up to numeric(18,X), but MsSql goes much higher (up to
28 and with a switch, up to 38). So, you'll need to add more "non-ANSI"
words, like int64 (or longlong) and if you make IB tomorrow to reach the
same precision than MsSql, would you invent longlonglong, for example?
After all, numeric(18,0) is not as bad. And "integer" still defines the same
integer you know, either in dialect 1 or 3. If you want to use all the range
of generators in dialect 3, then use numeric(18,0).
Let's keep in this way until the SQL committe invents a new funny name for
non-fractional types greater than integer, maybe sqlint64, who knows.
Just my opinion.

C.

> -----Original Message-----
> From: Dmitry Kuzmenko [mailto:dima@...]
> Sent: Jueves 10 de Agosto de 2000 10:46
> To: ib-architect@egroups.com
> Subject: [IB-Architect] Re: 32-bit integers
>
>
> Hello, All!
>
> > Please, answer me on one stupid question: Where are
> > 32-bit integers now in Interbase 6???
> > Who gave this crazy idea to stay with INTEGER keyword
> > but to change it to 64bit????????
>
> I'm apologizing about this really stupid message.
> I was mistaken with the numeric tests, and saw wrong results.
> Had a bad, bad day...
>
> Of course, INTEGER is still 32-bit in dialect 3.
> The only thing I don't understand is why INT64 datatype
> can't be used directly, but only via NUMECIC(18) (or like this).
> Can somebody explain why?
>
> --
> Dmitry Kuzmenko, Epsylon Technologies.
>
>
>
> To unsubscribe from this group, send an email to:
> IB-Architect-unsubscribe@onelist.com
>
>
>