A recent editorial in a major PC monthly magazine got me thinking about my days at IBM.
The editor lambasted Microsoft for its poor quality record and lack of true leadership and innovation. At one point, the editorial cited how much function had gotten thrown out of the upcoming Microsoft Vista OS and implied that this is a bad thing. Well, perhaps not. At least from my prior experience, here is how it worked when I was at Big Blue.
The term "development" applied to the groups, both hardware and software, that actually designed and developed the product. Of course, there were numerous other organizations involved - service, manufacturing, test, publications, etc. But development did not set the overall product direction; this was done by an organization called "planning". The job of the planners was to understand the business requirements and overall business strategy, and to refine those requirements into a set of objectives. Development then took those objectives and turned them into a product. The statement of what development was actually going to design was called the Functional Specification, which was usually written by development in response to planning input.
Very often, development just couldn't design a product to accomplish planning's goals, especially if time and cost were factored in. But the interchange, give-and-take, between planning and development was a very natural one as the planners really didn't understand the development restrictions, and development didn't always have enough visibility to the marketplace to determine requirements. Despite Scott Adams lampooning this in Dilbert (which I read with relish every day), the process actually worked fairly well.
But even after the Functional Specification was agreed to, there would often be last minute changes. Sometimes some function had to be delayed to later releases; this often occurred if there were unforeseen difficulties. But the process of finally agreeing on the product to be shipped was a natural and positive one. I was known for my skills in doing exactly that; I knew how to narrow a product down to what was needed and to concentrate on getting it "out the door".
One time, however, it did not proceed normally. We had a product that had a basic function for a certain price, and then more advanced function for a higher price. The higher price included more RAM memory to store the additional programming needed. One of our Very Senior Planners (VSPs we'll call them) decided it would be nice to let customers sample the advanced function even if they had the basic model. They called it (I kid you not) "Try it, you'll like it!" I went to see the VSP and patiently attempted to explain that we needed the extra RAM for the advanced functions and that it was impossible to put that much function into the basic product.
Well, the VSP first stared at me like I was something that crawled out of the slime, and then loftily explained to me that it was the job of development to implement what planning wanted, and that technical issues were development's problem. At which point he shooed me out of his office. We never found out the end of the argument as the product never actually was shipped in its basic form anyhow. Perhaps development won after all, perhaps not.
But the epilog to the story is that the VSP in question left IBM and took a very senior position at, you guessed it, Microsoft! So perhaps the comments of the editor in that magazine column were closer to the truth then he knew. Perhaps the Vista planning organization crammed it with so many diverse requirements that development just couldn't handle it, and some things had to be left for later releases. I think the difference between MS and IBM (at least in those days) was that IBM would not announce a product (and its functionality) too early, as MS does.
But in a few months (around the end of January we hear) we will all find out when Vista ships.
(Bob Seidel is a local computer consultant in the Southport - Oak Island area. You can visit his Website at www.bobseidel.com or e-mail questions or column ideas to him at bsc@bobseidel.com. For specific inquiries, please call Bob Seidel Consulting, LLC at 278-1007.)