This is more like a long Tweet than a short blog entry.

Personal issues aside Mark Hurd was a brilliant CEO for HP and he will be remembered as one of the greatest CEOs HP ever had. Like the rest of the world, we were curious and watched him from the beginning. I knew he was off to a brilliant start when - within days of getting the job - he put together an internal group of PhD economists specialized in game theory who developed a system that saved HP billions of dollars in purchases.

Wikipedia defines game theory as "attempts to mathematically capture behavior in strategic situations, or games, in which an individual's success in making choices depends on the choices of others."

Hurd got up early every morning watching Armonk like a hawk.

Steve Mills via satellite with Jonathan Schwartz during the opening keynote of JavaOne. (Photo copyright: Fuat Kircaali - Ulitzer, Inc.)

I worked with Steve Mills in 1985-86 at IBM's IS & CG headquarters at 44 South Broadway in While Plains, NY, reporting to Ralph Janegie, Jerry Fensterstock, and Curt Sanders.

Steve's IBM bio says "He became Director of Planning in the Information Systems and Communications Group in 1985. He was responsible for business planning and strategy for IBM's personal computer and communications products. In 1986, he was one of the executives responsible for starting IBM's Publishing Systems Business Unit. He became Director of Financial Planning at Corporate Headquarters in 1988."

I worked in his "printers and peripherals" group as a market research analyst. We used to massage data using a mainframe language called APL which came with its own APL keyboard.

My main responsibility was to compile a weekly management report on U.S. retail sales for printers. Twenty-five years ago most printers were still dot matrix differentiated only by narrow and wide paper types. Epson and Okidata were top brands. Our raw sales data came from Dataquest and a few other research companies and estimated national printer sales by brand and model based on same-store data. The information was pretty accurate. With our independent research, we probably knew the exact weekly sales for most printers and models before the manufacturers. I'm guessing a lot of the research we did went into the launch of Lexmark in 1991.

I was 25-26 years old when I worked at IBM. Steve was not much older. IBM, like many companies, plots a career path for every new employee it hires. It moves them from job to job grooming them from their first day on the job until their retirement. Whenever I took an elevator with Mills, people would whisper behind his back that he would one day be CEO of IBM. That was the rumor that followed Mills everywhere he went. Well, we heard the same rumor about Mike Armstrong, another big IBM name at 44 South Broadway. Armstrong was Steve's boss then but that next job didn't come through for him after successfully running IBM Europe. So he left and became the CEO of AT&T for a while, but that's a different story.

In a short Tweet the other day I said "Steve Mills will not take the job if offered, unless of course #IBM passes on him like it did with Mike Armstrong. #HP #IBM #MarkHurd #Cloud  7:28 PM Aug 16th  via web"

I haven't spoken to Mills recently and I don't know what he will or won't do. I know he is one of the "very few" qualified people to run IBM or HP. I hope he gets the top job at IBM one day. I also know he will finish his term squeaky clean.

• IBM Snags Netezza for $1.7BN