Re: Date format RR limitations. (adding; Please answer, Thomas Kyte!)

From: Remco Blaakmeer <remco_at_rd31-144.quicknet.nl>
Date: 2 Mar 1999 22:10:33 GMT
Message-ID: <7bhnkp$q6e$1@rd31-144.quicknet.nl>

In article <w1YC2.291$rS5.2036_at_news1.online.no>,

"Roy Brokvam" <roy.brokvam_at_conax.com> writes:
>
> Mike Burden wrote in message <36DBC595.8B97822C_at_capgemini.co.uk>...
>
> Help!
>
> I have assumed the same! As I interpret the SQL Language Reference Manual,
> you are quite right!
>
> For any current year between 1950 and 2049, any 2-digit year entered through
> RR will translate to a year between 1950 and 2049.
>
> If I repeat the above sentence, subtract 50 years and use YY instead of RR,
> it reads:
>
> For any current year between 1900 and 1999, any 2-digit year entered through
> YY will translate to a year between 1900 and 1999.
>
> Which is why YY is bad!

You could, of course, change all applications back to YY format around 2030 or 2040 or so, if you are sure that by then your database contains no dates older than Jan 1st, 2000.

The same goes, of course, for using the RR format now. If your database contains data older than Jan 1st, 1950, you should not use the RR format.

The YYYY format is just generally better. My company switched to DD-MM-YYYY format a year ago and as far as I know nobody really complained. They just felt like "wow, our applications are Y2K compliant already". :-)

Remco
--
rd31-144: 11:00pm up 2 days, 4:50, 5 users, load average: 1.07, 1.02, 1.00 Received on Tue Mar 02 1999 - 16:10:33 CST