The history of computing can be described as waves.
The first waves were purely of invention of basic ideas, where cogs gave way to vacuum tubes which gave way to transistors, and magnetic core memory gave way to dynamic RAM.
The next waves created scale, where individual components gave way to integrated circuits and silicon chips.
Then came additional waves of standardization where frames were used to allow multiple systems to interconnect, followed by the seven-layer models of networking.
Then came several waves of abstraction with hypervisors, virtual environments, distributed computing systems and structured programming languages with libraries.
Today we see wave after wave of new computing being driven my such ideas as Machine Learning based A.I., edge computing, IOT, Augmented Reality, Block chains and mobile computing and today it’s all backed by incredibly sophisticated sales and marketing.
A modern CIO has a lot to consider, and when a new business idea is presented and must be supported by ever more complex IT environments, the pressure is always on to deliver. And if you can base a new project on only new technology things become easier, partly because trying to deal with legacy systems adds complexity, but mainly because the people who run legacy systems have different ways of considering risk, and these ideas always make new projects seem a lot more complex.
When you visit with vendors trying to sell you shiny new tech, that will do amazing things, they are also under pressure to “close deals” and would much prefer to avoid the complexity and delays of trying to include all that “legacy stuff” in the deal.
But here’s the issues with this:
- Much of the worlds commercial data is already on legacy systems, and a huge amount of that is mainframe. Everything from financial records to medical information to scientific and government data is being securely and reliably hosted in ways that are prohibitively expensive and risky to change.
- The people that have been running legacy systems for many decades have been through many failed development and delivery cycles, and can spot problems in project plans, because they have already made the same mistakes in the past. When they tell you that a project will take eight times longer than projected and cost five times as much as projected, it’s because they’ve been battle hardened. It’s painful to hear, but these are not “old people complaining”, these are seasoned veterans who know exactly how to deliver high performance, highly scalable, highly secure, highly reliable business applications, because that’s exactly what they are delivering right now.
You may not always like to hear the message that your legacy teams have to deliver, and you may not always agree with them, or even accept their input, but they are where you will be soon enough.
What is new today, will be legacy next year.