We use cookies and other similar technologies (Cookies) to enhance your experience and to provide you with relevant content and ads. By using our website, you are agreeing to the use of Cookies. You can change your settings at any time. Cookie Policy.

Metadata: Thursday’s Child

Originally published October 21, 2004

Metadata from the beginning has been Thursday’s child.  Like the late Rodney Dangerfield, metadata gets no respect. If you ask any organization if they have a metadata initiative, you will find it will always – ALWAYS – be a low priority project. Yet metadata is important. It is very important. Something is amiss for there to be such a grievous disconnect. 

In order to understand why metadata has been Thursday’s child, let’s look at the past. Ask a programmer. Any self-respecting programmers will tell you their jobs are: “We are problem solvers. We write code and when the code is complete the problem is solved”. Where does metadata fit into this equation? It is as far removed from the mindset of the programmers as possible. The programmers look upon metadata as a concept: 

  • to be ignored;
  • that presents an obstacle to getting to the solution of the problem; and
  • a necessary evil to placate the powers that be. 

In short, metadata must be dealt with in the quest to solve problems as efficiently as possible. 

There is also the maintenance programmer. The maintenance programmer looks at metadata as a task to be avoided. Once the problem is solved, the maintenance programmer moves on to the next maintenance project. There is no consideration of maintenance for metadata. The problems of gaining respect start with the programmer. But the programmer and the programmer’s attitude is only the start of the problem. 

Review the classical technology solution used to “solve” the metadata problem. Look at data dictionaries and their repositories. Data dictionaries are a grand idea until you need to implement one. Eventually you find out that the work required to load data into the dictionary is burdensome. First, there is much metadata that has to be loaded. Second, the people that really understand the data retired or died last year. Third, the metadata changes as fast (or faster) than it can be loaded. Fourth, once loaded, the metadata is “optional” as to whether it can be used. No wonder data dictionaries and their second generation, central repositories, failed miserably. They were a great idea until they became a reality. Then they were failures. 

But there were other obstacles for metadata. Most of what passes for metadata management technology today, has its roots back in the days of the mainframe. In the day of the mainframe, all processing was centered on the mainframe. If processing was being done, it would be done on a central processor. Therefore, metadata needed to be managed centrally. 

But the world has changed. Today’s corporate information factory is a distributed world. There are end-users. There are departments. There are data marts. There are exploration warehouses. In short, the world of processing today is a diverse, distributed world. Trying to plug a centralized metadata management tool into a distributed world is like trying to put a round peg in a square hole. It just doesn’t fit. 

But there is one last problem with metadata that is perhaps the most important. It is that the end-user controls the metadata. When you stop and think of it, what is a spreadsheet? The columns and rows of a spreadsheet are nothing but metadata. And who controls the spreadsheet? Why the end-user does, of course. Tell an end-user that permission is needed to add or modify the columns and formulae found on a spreadsheet and see how far you get. The truth is, the whole reason the end-user got a spreadsheet in the first place was to control the content. The end-user is not about to let the programmer have ANY control over the metadata that goes into the spreadsheet. The programmer can create the most powerful and elegant metadata repository but will not have the slightest effect on the end-user when it comes to doing analysis.  

There are undoubtedly many reasons why metadata has far to go and been pushed to the side. These are just some of the most obvious reasons. But is the world of metadata to remain is such a miserable and hopeless state forever? The answer is NO! The future for metadata is changing.

  • Bill InmonBill Inmon

    Bill is universally recognized as the father of the data warehouse. He has more than 36 years of database technology management experience and data warehouse design expertise. He has published more than 40 books and 1,000 articles on data warehousing and data management, and his books have been translated into nine languages. He is known globally for his data warehouse development seminars and has been a keynote speaker for many major computing associations.

    Editor's Note: More articles, resources and events are available in Bill's BeyeNETWORK Expert Channel. Be sure to visit today!

Recent articles by Bill Inmon



Want to post a comment? Login or become a member today!

Be the first to comment!