Comments on: Venn Diagrams explain SQL Joins Perfectly http://programmer.brettveenstra.com/2007/10/12/venn-diagrams-explain-sql-joins-perfectly/ One Programmer, on the Journey Tue, 07 Feb 2012 17:16:23 +0000 hourly 1 http://wordpress.org/?v=3.3.1 By: mikhael http://programmer.brettveenstra.com/2007/10/12/venn-diagrams-explain-sql-joins-perfectly/comment-page-1/#comment-1288 mikhael Tue, 12 Oct 2010 20:40:03 +0000 http://journeymantomaster.wordpress.com/2007/10/12/venn-diagrams-explain-sql-joins-perfectly/#comment-1288 <p>That's one way to get at what's going on, but if you loose focus on your original sets, the original digram has failed to describe your JOIN and you have to come up with another. So I guess "perfect" is not the word I would choose.</p> <p>If two tables are in a one-to-one relationship with each other, the original two-circle diagrams happily describe a given join. It's when you introduce 1-to-many relationships that we are forced to create a new set that neither of the original circles represent even partially. In this scenario we are not simply identifying singular objects that share a given characteristic, we are creating new objects/records that fit our description that did not exist previously.</p> <p>In other words, when we say "Give me the LEFT OUTER JOIN" between two tables that have a one-to-many relationship, we are saying "The Records from the LEFT table are not good enough to satisfy what I want. The records in the Right table are not what I want either. Instead give me something completely different based on these Left and Right components."</p> That’s one way to get at what’s going on, but if you loose focus on your original sets, the original digram has failed to describe your JOIN and you have to come up with another. So I guess “perfect” is not the word I would choose.

If two tables are in a one-to-one relationship with each other, the original two-circle diagrams happily describe a given join. It’s when you introduce 1-to-many relationships that we are forced to create a new set that neither of the original circles represent even partially. In this scenario we are not simply identifying singular objects that share a given characteristic, we are creating new objects/records that fit our description that did not exist previously.

In other words, when we say “Give me the LEFT OUTER JOIN” between two tables that have a one-to-many relationship, we are saying “The Records from the LEFT table are not good enough to satisfy what I want. The records in the Right table are not what I want either. Instead give me something completely different based on these Left and Right components.”

]]>