Re: The C in ACID
Date: Mon, 21 Aug 2006 18:02:10 GMT
Message-ID: <CymGg.433797$Mn5.256414_at_pd7tw3no>
Paul Mansour wrote:
> paul c wrote:
>
>> I don't see a conflict. Perhaps you see a subtlety I don't. I consider >> the facts in a db correct if they are consistent with whatever >> constraints I've specified. If a user enters an untruth, there's nothing >> the dbms can do about it, unless it has been told, eg., with >> constraints, how to recognize certain lies.
>
> I was under the impression the terms Correct and Consistent had
> precise, well-defined and even generally agreed upon definitions with
> respect to the relational model. For example, in CJ Date's 8th edition:
>
> "Correct implies consistent (but not the other way around) and
> inconsistent implies incorrect (but not the other way around) where by
> correct we mean the database is correct if and only if it fully
> reflects the true state of affairs in the real world"
> ...
Being very literal about "Correct implies consistent", a db could be incorrect and still be consistent. Whereas "database is correct [IFF] if it fully reflects [TRUTH]" says something different. I don't see how a db can record truth, only how we interpret its answers as true.
> Obviously my impression was wrong.
>
> It does seem a little strange that Lomet and Barga would have the C in
> ACID mean correctness, (even if they call it consistency), and then say
> that the DBMS can't enforce it. Why bother having C in the acronym if
> it can't be enforced?
>
I vaguely recall numerous earlier sources that intended the "C" to stand for Consistency.
p Received on Mon Aug 21 2006 - 20:02:10 CEST