"Alan" <alan_at_erols.com> wrote in message
news:b717p3$a176q$1_at_ID-114862.news.dfncis.de...
> "--CELKO--" <71062.1056_at_compuserve.com> wrote in message
> news:c0d87ec0.0304081511.36fdbd40_at_posting.google.com...
> > >> Let's say you have a table of product orders and you need to know
> > which one's have been shipped and which one's haven't. That's a
> > boolean value that can't be computed from other data. <<
> >
> > Wrong:
> >
> > CREATE TABLE Orders
> > (order_nbr INTEGER NOT NULL PRIMARY KEY,
> > ..
> > ship_date DATE, -- null means not yet shipped
> > ..);
> >
> > And please don't tell that you would not record the shipping date,
> > shipping charges, shipping method, etc. in the database.
>
> So, you are saying NULL may have a valid semantic meaning? Doesn't NULL
(in
> this example) mean, "have not a clue, maybe it was shipped, maybe it
wasn't,
> maybe it will be shipped one day..."? Shouldn't it work like this:
>
> IF Shipping_Date <= Today, then it's been shipped
> Else If Shipping_Date > Today, it has not yet been shipped, but apparently
> will be on the given date
> Else If Shipping_Date IS NULL, then we just don't know...
What happens if something goes wrong and the item doesn't ship on the
anticipated date? Hmmmm.... I guess you might need a boolean, after all, to
indicate whether the item actually shipped.
Received on Wed Apr 09 2003 - 11:40:12 CDT
