Michael Stonebraker Explains Oracle’s Obsolescence, Facebook’s Enormous Challenge
After reading this article, one thing I found was that Stonebraker gave much of the credit to what Codd said. Michael Stonebraker brought relational techniques to companies that produces billions of dollars in sales. He acknowledged that what allowed him to do what he did was simply taking the relational technology talked about by Codd and putting it into existence. Codd saw that we should not build specific to the data. This can cause problems because data can be around for a very long time. For example, a sporting goods company might grow and move into the outdoor space as well. This will cause for the data structures to change. If the company is using the assembly processes that was common when Codd came out with his work, you would have to start over from scratch whenever new data is introduced. However, if you were to code at a higher level that has data independence, then new data will not cause these problems.
Another thing I found was that Stonebraker says Facebook has this huge graph problem. The problem is they try to find the average distance from one person to another. Basically, I think it means that Facebook's answer to the problem is one that MySQL cannot solve which is the program used by Facebook. So, they are trying to find a solution that would be good enough to be commercialized.
A final thing I found was the NoSQL is doing almost the opposite: instead of creating higher level languages it creates lower level languages. This is interesting because they are hoping that the higher level languages do not work out and it falls back to their lower level languages. Stronebraker believes that it keeps getting closer and closer to being like SQL and eventually they will become a high level language in the same market as SQL.