From: --CELKO-- on 2 Dec 2009 17:37 Is the schema really full of redundant tables with improperly named columns? When you see tables with identical structure, you know that you have attribute splitting. We use see the magical generic "id" of nothing in particular. How can you have a NULL user? In fact, this has more NULL-able columns than most major corporate databases I have ever seen. At best, you are actually writing 1950's style mag tape merges with UNIONs mapped into SQL and not an RDBMS at all. This is why people are going nuts trying to figure out what your specs are. All you will get with DDL this bad is a stinking dirty kludge. >> I am trying to produce a single select statement which is true for two separate cases (so I can reuse it within an IN expression.. which means no IF statements).<< Look up the concept of COHESION in any book on software engineering. A module of code should do one and only one function -- not two, not twenty. This has nothing to do with SQL. This is absolute foundations of programming and you don't know it. We need to start over and do it right.
First
|
Prev
|
Pages: 1 2 Prev: Insert trigger Next: xml.value xquery on element centric xml parameter |