IT Management and Cloud Blog

« | Home | »

What does Open Source Mean?

By John | September 5, 2007

In the late 1980s, IBM made a bold move and turned all its code to OCO (object code only). Until that point, all IBM’s operating systems and transactions software (e.g., IMS or CICS) were available as source code via microfiche. My first job was as a systems programmer at Exxon’s geophysical processing division, which searched for oil by looking at seismic data. These were the scientists who founded the Prudhoe Bay oil field, an area that in 1988 was producing around 2 million barrels of oil a day during what some might remember as the oil rush of the ’80s. These guys could not and would not be restricted to silly things like 32k record sizes and memory segments that couldn’t exceed 16 meg, so we modified the operating system kernel anywhere we could.


At one point, we had over 250 modifications to the base IBM operating system startup modules. The beauty of the situation was that Exxon could never have handled the streams of seismic data that we were processing without our modifications. In addition to this cumbersome dependence created by OCO, IBM also suffered a diminished quality when it went to OCO. Back in the day, when you had a problem with an IBM program, you could pull up the source offset and see exactly what instruction was being executed. IBM had thousands of debuggers, and, usually when a problem record was opened, the solution was also provided by the customer. So, when I hear the phrase open source, I guess I am old school. It means, to me, that I can see the source code, and that’s a really good start.

Topics: OSS, esm, ibm, opensource | No Comments »