Re: forms30 and Y2K

From: Steve Cosner <stevec_at_zimmer.csufresno.edu>
Date: 31 Jul 1998 15:58:58 GMT
Message-ID: <6pspk2$e2q2_at_hendrix.csufresno.edu>


In article <6pqqsk$7s3$1_at_nnrp1.dejanews.com>, <ironmtn_at_my-dejanews.com> wrote:
>I have been reading the available newsgroup information and Oracle metalink
>postings and have a question.
>
> If you set nls_date to mm.dd.rrrr (or rr), when the database
> receives dates from forms in the format of mm/dd/yy, shouldn't it
> force the correct century at storage time?

No!

> The only reason I can think of as to why this wouldn't happen is if
> the forms tool itself passes a century to the database that is
> incorrect (because forms30 doesn't have rr capability).

That's right. The only time the nls_date format and the RR century indicator has any effect is when you pass a text string to the server using the to_date function without a format mask. Then the nls_date_format becomes the default mask. Forms passes dates to the server as date datatype, so the century is already established by Forms. The RR date format cannot be used in Forms 3.0, if I remember correctly.

I have a "sliding century" function that I used in Forms 3.0 (and now in Forms 4.5) to infer the correct century. -- it's better than RR, because it is date sensitive, and we use it almost everywhere in forms. The biggest problem is that you need to call it from every date validation trigger.

I can email it or post it if anyone wants to see it.

Steve Cosner Received on Fri Jul 31 1998 - 17:58:58 CEST

Original text of this message