Re: Alternatives to ACID, non-serialized concurrent transactions?

From: Paul Vernon <paul.vernon_at_ukk.ibmm.comm>
Date: Wed, 7 May 2003 17:34:20 +0100
Message-ID: <b9be8o$1vs0$4@gazette.almaden.ibm.com>

"tj bandrowsky" <anakin_at_unitedsoftworks.com> wrote in message news:af3d9224.0305061537.137f3667_at_posting.google.com... [snip]
> Now, forcing a database to have a value at one point in time is
> entirely valuable, otherwise, we wouldn't all be doing it.

We are all doing it so it must be right? In this case what we are all doing is right, but beware, your line of argument can be a logical fallacy.

> What I'm
> asking if there is not some framework that is not actually more
> valuable.

OK, that is potentially a fair enough question. Just not one that I think is an interesting one.

However, it occurs to me that a 'logically distributed database' could be exactly such a framework. I would imagine that one critical logical difference between a 'logically distributed database' and a 'logically single database', is the fact that the former would not necessarily have one value at every given point in time.

Regards
Paul Vernon
Business Intelligence, IBM Global Services Received on Wed May 07 2003 - 11:34:20 CDT