Oracle FAQ Your Portal to the Oracle Knowledge Grid
HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US
 

Home -> Community -> Usenet -> comp.databases.theory -> Re: What is this model technique called

Re: What is this model technique called

From: --CELKO-- <71062.1056_at_compuserve.com>
Date: 7 Jun 2003 18:48:36 -0700
Message-ID: <c0d87ec0.0306071748.25cc57c6@posting.google.com>


>> ...; Then for those entities where we believe there will be
"spill-over" attributes, we plug-in the EAV architecture. <<

What the heck is a "spill-over attribute"? This sounds like you mean "a bad design" and you ned to do some research. \

>> Then, after a few years into production, when the new attributes
have
"matured" into first class citizens, we add a new field [sic] to said table or
introduce 1:1 subtype with the additional fields [sic] ? <<

These are attributes, not wine. And columns are not fields. If the design needs to change, you go back to the ORM or other high level conceptual model. You re-create the schema from that. Then you develop an ETL tool that can load the old data into the new schema. What you propose is like patching code until it cannot be maintained -- only databases fall apart faster because some many applications use them. Received on Sat Jun 07 2003 - 20:48:36 CDT

Original text of this message

HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US