Re: Nested structures

Bob Badour wrote:
> Rob wrote:
>
> > Where can I find this paper?
> >
> > dawn wrote:
> >
> >>Some, but not all, of you will find this new IDC paper entitled
> >>"Because Not All Data is Flat: IBM's U2 Extended Relational DBMSs" to
> >>be of interest.
> >>
> >>Other than the fact that the term "flat" is used, which I know can be
> >>inflammatory in this NG, are there other logical responses?
> >>
> >>I think the term "embedded" is misleading, given the number of
> >>companies that use these DBMS's for their enterprise systems. But I do
> >>understand that IBM has to be careful to position each of their
> >>offerings somewhat differently.
> >>
> >>The information about the embedded market growing faster than the
> >>relational market is quite interesting. Perhaps there is and should be
> >>a different theory related to those DBMS's currently called "embedded"?
>
> Non-relational products are nothing new for IBM. They have traditionally
> helped IBM sell mainframes.

These are relatively new, having been acquired by IBM when they purchased the assets of Informix. Few at IBM seemed to have heard of the U2 databases at the time, so they were surprised to learn they were actually a customer.

> There are already lots of embedded SQL dbmses. It should come as no
> surprise that a proper superset is growing faster than its proper
> subset--it would take some doing to grow slower. Yawn.

Many, but not all, of the databases marketed as embedded are not "traditional/legacy SQL databases" in that they do not require the developer to put data into first normal form (while higher level normal forms, sans their 1NF requirement, are still among best practices for modeling data for these products).

--dawn Received on Fri Sep 15 2006 - 23:56:29 CEST