How Many Software Developers Does It Take to Screw in a Light Bulb?


The summer solstice/Fourth of July timeframe is my New Years. I mean it's the time when I buy into the concepts traditionally hyped around Jan. 1 in cultures influenced by the Roman Empire. But it's also new year, as in the year of the ox, Rosh Hashanah, the first Sunday of Advent, etc. all rolled into one. I guess this is some throwback to the school year (which in my case was a long, long time ago) or more likely to the way the two major market-research firms I worked for ran their businesses (determining at the beginning of the Northern Hemisphere summer what they would "sell" in the following calendar year).


So this is the time to start thinking about what's next in my IT research agenda, to level set, to get off the tangents down which breaking news or personal IT likes or dislikes have led my opinionating, to go back and confirm or deny assumptions and definitions that affect the answers you provide me in my research, to question all my own IT biases. (If you have responsibility for an IT budget that runs on a calendar year, you might want to think the same way about that process in this timeframe. I think the bean counters call if zero-based budgeting.)


If I hadn't needed my own background and experience to start this process off, Oracle did it for me by beginning to deliver on July 1 the "middleware" upgrade it had promised on May 5, 2007. At various times in July and August I'll take a look at what's new in that Oracle announcement and provide some opinion. The mass of the Oracle message makes almost any one-paragraph summary you've seen to date in the blogs or technical press frivolous.


But in a sense the Oracle "middleware" announcement also sets my 2009-2010 research agenda. I see five overarching questions in the Oracle announcement that are the most important issues you face as an IT staffer or manager, not just in 2010 but throughout the 2010-2019 decade:


  1. What is middleware? This is important because in defining middleware we define the entire IT stack, what you are buying, and what its value proposition is.
  2. What is an IT stack? If it means everything from the chip set to the bill of material processor, instead of just deployment software, then doesn't that also mean we're going back to time-sharing service bureaus (only we call it cloud computing now)?
  3. What is the best way for you and your IT supplier to fund/pay for the development of all of the above in whatever manifestation (point product, stack, platform as a service) results? To many of you it is clearly no longer the two-tiered license/maintenance fees revenue model that Oracle kind of invented around 1980. So what's next? And remember there is no such thing as a free lunch, even on New Year's day.
  4. What is the best way to do the development, no matter how it is funded? Is it better to amalgamate and integrate existing commercial off-the-shelf technologies the way Oracle has been doing since its May 5, 2007 middleware roadmap announcement, or is it better to keep re-inventing the wheel via in-house and/or custom development efforts? How does the open source software culture and open source license terms and conditions (and patents and copyrights and all kinds of legal issues) affect this funding? How many enterprise software developers will it take to install a light bulb?
  5. What are Google, IBM, Microsoft, Oracle and SAP going to do in the next 10 years from a business and revenue model point of view in response to the likely answers to the above four questions? (The list is alphabetic but these are the top five software suppliers worldwide as measured by IT Investment Research methodology, which includes more than just license/maintenance revenue.) If your preferred IT vendor is not on this list, how should you alter your plans?


As you start planning for the next decade or even the next calendar year, what are your cosmic questions? Send me an e-mail or post a comment.