From: "Scot A. Becker" Subject: Re: Database design Date: 2000/01/01 Message-ID: #1/1 References: <385DE338.6488F4AA@wizard.ae.krakow.pl> <386CE067.95E227C6@concord.com> X-Priority: 3 X-MimeOLE: Produced By Microsoft MimeOLE V5.00.2919.6600 X-Complaints-To: abuse@mediaone.net X-Trace: typhoon3.rdc-detw.rr.com 946692443 209.32.180.171 (Fri, 31 Dec 1999 20:07:23 CST) Organization: InConcept, Inc. X-MSMail-Priority: Normal Reply-To: "Scot A. Becker" NNTP-Posting-Date: Fri, 31 Dec 1999 20:07:23 CST Newsgroups: comp.databases "Robert Kolker" wrote in message news:386CE067.95E227C6@concord.com... > You can't do relational database design without knowing something about normalization. Being a bit of a nitpick , this isn't quite right. Using Object-Role Modeling, one doesn't need to know anything about formal normalization. ORM's (elementary) fact based approach insures that the resulting logical (ER) schema is in "optimal" normal form, which, in practice, is at least 4NF, and often 5NF (because ORM captures more rules than ER approaches, many 5NF errors are also avoided). I've written a bit more about this topic at the URLs below. Specifically, the articles can be found at http://www.inconcept.com/JCM/June1999/becker.html and http://www.inconcept.com/JCM/August1998/becker.html. HTH, Scot. ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Scot A. Becker Principal Consultant, InConcept, Inc. http://www.inconcept.com Editor, The Journal of Conceptual Modeling http://www.inconcept.com/JCM