| Oracle FAQ | Your Portal to the Oracle Knowledge Grid | |
Home -> Community -> Usenet -> comp.databases.theory -> Re: Named Mistakes and Questionable Practices
David Fetter wrote:
> ...
>> 4) Denormalized tables
There are occasions when I'd say it's not idiotic at all. Updating is another matter.
I think any decent catalogue ought to enable some kind of comment for users, not just developers.
>> 5) Auto-increment methods for keys
I think the proper objection is against wanton/willy-nilly use of generated keys. Some accountants prefer this so that no entry is cancelled. Time-of-day is often generated and seems quite 'natural' to me even if it isn't a user who enters it. Sometimes their use is a signal of no effort spent discovering the real business keys. One danger in such a prescription is that you see one example and conclude that the rest of the system is bad, too. My advice when assessing such dictums is look at how the people who advocate make their money and ask whether the advice serves them or the user.
One problem with these lists is that people get dogmatic about them, like some "rules-of-thumb" I've seen. Excuses for not thinking about the problem at hand. From the user's or customer's point of view they can also produce systems that are way more than what they want/need, with tons of lookup tables that have to be maintained every time some standards body revises its collection, although this does have the developer advantage of jobs for the boys. Some businesses could care less about SIC's or ISO country codes. Maybe those aren't so bad, but the codes in the airline biz are a joke.
Personally, I wouldn't look to an 'industry standard code list' unless it was an explicit biz requirement, specified by the user/customer.
That's my rant for today!
p Received on Mon Jun 12 2006 - 19:25:55 CDT
![]() |
![]() |