Metadata is data about data, but what is metaprocess information? Metaprocess information is data about process. In most ways, metaprocess information is the yen to metadata’s yang. They are as complementary as black and white. As left and right. As night and day.

While everyone has heard of metadata, almost no one has heard of metaprocess information. Metaprocess information is information about process. What might information about process look like? Some of the more interesting aspects of metaprocess information include:

• The process name or other identifier of the program or process
  
  
• The input to the process
  
  
• The output from the process
  
  
• The variables used in calculation and algorithmic processing by the process
  

If you happen to recall the old HIPO charts, the metaprocess information mirrors very much what a HIPO chart is all about. This short list merely scratches the surface as to what the actual contents of metaprocess information might be.

But what good is metaprocess information? If you have a bunch of old legacy systems and you need to document those old systems, then you might find metaprocess information quite useful. Suppose you have a pile of 1,000 old programs. Suppose you want to migrate to a new system. Or suppose you want to merge with another company. Or suppose you just want to do plain old maintenance. There are a thousand reasons why you might need to understand what is in your old legacy systems.

Now if you have a lot of those systems, trying to churn through line after line of old source code is beyond tedious. Churning through mountains of old source code surpasses the capacity of even the brightest minds. At some point in time, even Einstein would lose the ability to keep track of all the details of what is going on in old legacy systems.

So you have a problem. You need to understand what those old legacy systems do, but there is not enough manpower in the world to intellectually attack the inner workings of 1,000 old source code programs. What you need is metaprocess information that has been generated automatically. If somehow you could produce that metaprocess information automatically, you could address the mess that your old legacy systems are in today.

What is required to address metaprocess information in an automated manner? There are a variety of challenges.

The first challenge is that metaprocess information all begins with source code, i.e., text. Say the word “text” to most technologies and they immediately start to melt (sort of like Superman and kryptonite.) Most technologies have only the vaguest idea as to how to handle text. So text is the first challenge in getting your hands on metaprocess information.

The second challenge is that in order to create metaprocess information, the program that automatically generates the metaprocess information must be able to understand the logic found in the source code on which it operates. At the minimum, the source code’s logic must be followed (at least to a small extent) in order to determine what the appropriate metaprocess information actually is that is buried in the legacy source code.

The third challenge is that there are a myriad of reasons why old legacy code is difficult to handle. Some old legacy code is out of date or lost. Other legacy code works with one version of the base technology but not another version. Yet other source code has been physically degraded over time. In a word, legacy source code is usually in a pitiful state, where even physically reading the code may be a challenge.

There are obviously formidable challenges when it comes to unlocking the metaprocess information that is buried in your old legacy source code. But the value of being able to document old legacy environments in an automated manner is such that those obstacles will be overcome. It is like telling someone in Florida that gold will never be recovered from the sunken Spanish galleons that lie beneath the water and sand off the shore. The Spanish treasure is covered by sand and twenty feet of water. There are sharks and other deadly creatures that live in the ocean, but the lure of gold and treasure is such that even these formidable obstacles will be overcome. The same is true for metaprocess information.

SOURCE: What is Metaprocess Information?

Recent articles by Bill Inmon

• Deriving Business Value from Unstructured Nonrepetitive Data
• Characteristics of Corporate Data Types
• Business Relevancy of Corporate Data
• Classification of Corporate Data