“You can please some of the people all of the time; you can please all of the people some of the time; but you can never please all of the people all of the time.”
I like the way you have described the history of DW. Specially the phase I believe we are in now. All the vendors throwing platforms and tools at us but the tools or the platforms were never the "main" problem for us in DW-industry. It’s also funny to me (and sad at the same time): When attending both Hans Hultgrens and Dan Linstedts DV-courses, we are taught that both 3rd normal form or Dimensional Modeling approaches are bad and therefore choose DV..and then we end up using dimensional model to represent the DataMarts in a DV-DW. We end up having all the issues and benefits from DV and dimensional modeling in the same DW!? So practically we haven’t solved anything because to be flexible/scalable in DW-integration we used DV (and it’s not user friendly) and to make it user friendly we used Dimension modeling (which is not flexible). Ending up with 2 DW:s but we call in one DW. Then people in DV community will say, you don’t have to create a physical dimension model layer, just use view layer. To these people I say: It probably works in small DW:s but I have only worked with large enterprise DW (banking and telecom) and using views doesn’t really work when the users wants to analyze 10 years of historical data (transactions and call records)
I have been following your work and I saw in one of the videos that you need a real case hook-implementation. I am actually working with that now. My plan is to combine hook method and combine it with Puppini bridges. In this way we kind of combine best of both worlds. I just have to figure out how to “connect” hook keys with the concept of puppini bridges keys if you knw what I mean. I feel that something beautiful can be done here (combining hook and puppini keys)…just haven’t had the time to figure it out…yet.
That is great news. I'm so happy that you can see the benefits of the HOOK approach. We are using it on the project I'm working on right now, and it's only when you do things for real that all the little problems pop up, but so far it has gone really well.
And I can't quite believe that you are also looking at the Puppini bridge, as this is something I've also been investigating recently. In fact, I should be speaking with Francesco tomorrow to discuss some challenges I've been having. I'm more than happy to share my findings.
If you need any, ANY, assistance with your project then please feel free to ask. It seems we are trying to solve the same kind of problems and I'm sure we can help each other.
I have a quite a few more chapter to share but most will need some major editing. I'm hoping interest will begin to snowball, so please feel free to share.
I like the way you have described the history of DW. Specially the phase I believe we are in now. All the vendors throwing platforms and tools at us but the tools or the platforms were never the "main" problem for us in DW-industry. It’s also funny to me (and sad at the same time): When attending both Hans Hultgrens and Dan Linstedts DV-courses, we are taught that both 3rd normal form or Dimensional Modeling approaches are bad and therefore choose DV..and then we end up using dimensional model to represent the DataMarts in a DV-DW. We end up having all the issues and benefits from DV and dimensional modeling in the same DW!? So practically we haven’t solved anything because to be flexible/scalable in DW-integration we used DV (and it’s not user friendly) and to make it user friendly we used Dimension modeling (which is not flexible). Ending up with 2 DW:s but we call in one DW. Then people in DV community will say, you don’t have to create a physical dimension model layer, just use view layer. To these people I say: It probably works in small DW:s but I have only worked with large enterprise DW (banking and telecom) and using views doesn’t really work when the users wants to analyze 10 years of historical data (transactions and call records)
I have been following your work and I saw in one of the videos that you need a real case hook-implementation. I am actually working with that now. My plan is to combine hook method and combine it with Puppini bridges. In this way we kind of combine best of both worlds. I just have to figure out how to “connect” hook keys with the concept of puppini bridges keys if you knw what I mean. I feel that something beautiful can be done here (combining hook and puppini keys)…just haven’t had the time to figure it out…yet.
I will keep you updated.
That is great news. I'm so happy that you can see the benefits of the HOOK approach. We are using it on the project I'm working on right now, and it's only when you do things for real that all the little problems pop up, but so far it has gone really well.
And I can't quite believe that you are also looking at the Puppini bridge, as this is something I've also been investigating recently. In fact, I should be speaking with Francesco tomorrow to discuss some challenges I've been having. I'm more than happy to share my findings.
If you need any, ANY, assistance with your project then please feel free to ask. It seems we are trying to solve the same kind of problems and I'm sure we can help each other.
Looking forward to reading the next chapters. I resonated a lot with your problem setting and observations. Great work, Andrew
Thanks for your support Carlos.
I have a quite a few more chapter to share but most will need some major editing. I'm hoping interest will begin to snowball, so please feel free to share.