Comments on: The core idea(s) of DBMS2 http://www.dbms2.com/2005/10/10/the-core-ideas-of-dbms2/ Choices in data management and analysis Fri, 25 Jul 2008 01:35:00 +0000 http://wordpress.org/?v=2.5.1 By: Curt Monash http://www.dbms2.com/2005/10/10/the-core-ideas-of-dbms2/#comment-9 Curt Monash Thu, 13 Oct 2005 07:38:46 +0000 http://www.dbms2.com/2005/10/10/the-core-ideas-of-dbms2/#comment-9 Hi, Warren. There's a clear TCO argument for data warehouse appliances, as opposed to centralized RDBMS. There's a clear TCO argument for non-relational architectures when managing text, security data, etc. Those are examples of the TCO argument AGAINST getting too carried away with relational data integration. I suspect from your comments that you're a pragmatist. Good for you. Some of my more strident arguments are directed at the idealists -- i.e., Chris Date and his disciples, who seem to argue that (almost) all comments based on experience with, or the capabilities of, relational DBMS are invalid, because commercial DBMS aren't REALLY relational, and their double-secret TransRelational(TM) DBMS will change everything, automagically making a lot of programming and data administration tasks go away. Please see some of what are, at this writing, my most recent blog posts, especially the one dated 10/13. Thanks, CAM Hi, Warren.

There’s a clear TCO argument for data warehouse appliances, as opposed to centralized RDBMS. There’s a clear TCO argument for non-relational architectures when managing text, security data, etc. Those are examples of the TCO argument AGAINST getting too carried away with relational data integration.

I suspect from your comments that you’re a pragmatist. Good for you. Some of my more strident arguments are directed at the idealists — i.e., Chris Date and his disciples, who seem to argue that (almost) all comments based on experience with, or the capabilities of, relational DBMS are invalid, because commercial DBMS aren’t REALLY relational, and their double-secret TransRelational(TM) DBMS will change everything, automagically making a lot of programming and data administration tasks go away.

Please see some of what are, at this writing, my most recent blog posts, especially the one dated 10/13.

Thanks,

CAM

]]> By: Warren http://www.dbms2.com/2005/10/10/the-core-ideas-of-dbms2/#comment-8 Warren Thu, 13 Oct 2005 01:45:19 +0000 http://www.dbms2.com/2005/10/10/the-core-ideas-of-dbms2/#comment-8 Curt, None of the 3 topics you mention show me any reason why a RDBMS is not valid. 1. Organizations don't have enterprise data models. So what? I live with it every day. Doesn't stop me in any way from designing and implementing relational databases to support business needs. This point really makes no sense. 2. See above. 3. TCO? Where is the TCO for your, as yet, undefined, undeveloped, non-existant RDBMS2? Why would another application of database technology be cheaper to implement than relational technology? Again, this point makes no sense. So, I see three points, but not one of them shows the reader why they should look at an alternative to relational technology. Could you explain in more detail? I'm sure there must be a valid point somewhere, I'm just not seeing it. Warren Curt,

None of the 3 topics you mention show me any reason why a RDBMS is not valid.

1. Organizations don’t have enterprise data models. So what? I live with it every day. Doesn’t stop me in any way from designing and implementing relational databases to support business needs. This point really makes no sense.

2. See above.

3. TCO? Where is the TCO for your, as yet, undefined, undeveloped, non-existant RDBMS2? Why would another application of database technology be cheaper to implement than relational technology? Again, this point makes no sense.

So, I see three points, but not one of them shows the reader why they should look at an alternative to relational technology. Could you explain in more detail? I’m sure there must be a valid point somewhere, I’m just not seeing it.

Warren

]]>