In Champions Of Change: The App Dev & Delivery Role In 2020, we began to think about the business climate in the year 2020 and how it affects the application development and delivery role. Building on that theme, turn your attention for a moment to your existing application portfolio.
Ugh! Yes, that dark, dank, ugly bucket into which you've been dumping applications, enhancements, and upgrades for decades. That place where even though it is overflowing, only a few intrepid souls have the courage to look.
What do you see? Duplication? Yes. Waste? Yes. Needless heterogeneity? Yes. A tangled mess of point-to-point, siloed, marginally integrated apps and data seething and roiling with cost, complexity and other innovation-crushing demons?
If you are both truthful and like most of your peers, there is only one answer: Yes, to all of the above! OK, so let's stipulate that's at least partially true for all of us and that there is a chasm between that "place where demons be" and where your business leaders would like to be today.
How will you begin to sort it out, state its health and future usefulness, then reshape it toward the future that awaits us in 2020? Do you even try?
Here are a few schools of thought meant to spur debate:
Scenario 1 - You don't even try, because you know you'll rewrite all those apps before 2020
May I just point out that your predecessors said the same thing about those 35-year-old COBOL programs still in your portfolio today?
Perhaps we'll have such powerful semantic modelling and generation in 2020 that apps will write themselves. However, we've been predicting that technology will no longer require developers since the 1980s... still waiting.
Scenario 2 - You have blind faith that between now and then, someone will develop a magic tool to decouple valuable business functions from monolithic apps
Blind faith - No, not the rock band composed of Eric Clapton, Stevie Winwood, Ginger Baker, and Rick Grech, but a simple belief that someone will write analytical software and combine it with Watson-like supercomputing power to decompose your monolithic apps so you can recompose them into elastic applications more appropriate for cloud and mobile architectures.
The tool will strip out the third generation housekeeping code from the truly valuable code, isolate it, publish an API for it and leave it for you to orchestrate and reuse in an elastic application re-engineering component of the tool. Hey, it could happen!
Skeptics will say that "Relativity" and "Seek" tried and failed to isolate "business rules" from legacy code and expose them as XML, COM and EJBs in the 1990s. They did, but will it stay that way? Have we really not advanced the science of smart analytics since then, and will no one apply modern techniques to this monumental problem?
Why did they name the band "Blind Faith"?
With such superstars as Eric Clapton and Ginger Baker from Cream and Stevie Winwood (Spencer Davis Group and Traffic), promoters believed that people would buy the first and only album this group ever produced based on "blind faith." Seems we need a similar powerhouse team to solve the legacy software analysis problem.
Scenario 3 - You get started and pray for a hybrid... you begin to assess and streamline your portfolio manually, hoping that tooling improves
OK, so apps portfolio management tools are lacking (as the research below points out), but that doesn't mean you can ignore the task! Isn't that how it got so ugly to begin with? Much of apps portfolio management is process, and it won't happen by itself.
Some sort of health assessment is only one of several tasks on your plate if you want to champion change (see the research on becoming a BT leader).
Scenario 4 - I'm throwing down the gauntlet. What's your scenario?
How will your existing portfolio morph and change as cloud, mobile and other disruptive technology and business events reshape our world?
Become engaged - participate, cogitate and learn something about your future!
Posted by Phil Murphy