The world’s oldest software still in use
Imagine this: The year is 1958. You are Head of Technology at the Defence Contract Management Agency (DCMA) in the US. You have just spent a substantial amount of time and budget developing a mainframe system to handle defence contracts. Now – imagine someone telling you that 58 years into the future this very same system will still be running, handling data worth $1.2 trillion. Will you believe them?
The ultimate return on investment
This mainframe was the Mechanization of Contract Administration Services (MOCAS). It’s still running today, which most likely makes it the world’s oldest currently used software, showing an incredible ROI. It started as a well-designed, Cobol-based system that did the job it was supposed to do, but eventually it created a critical dependency where it simply couldn’t be swapped out. Not without huge potential disruption. So it kept running, year after year, despite being increasingly inefficient.
How long does software last?
These days, we don’t expect software to last for 50 years. We have seen over the past two decades just how difficult it is to predict the progress of technology. But back in 1958, people expected software – and everything else for that matter – to last a lot longer than we do today.
We no longer buy clothes that we expect to hand down to our children. Instead we buy affordable, dispensable items that are fashionable this season, knowing they probably won’t be trendy next year. And often we treat software the same way.
Situations like the MOCAS still occur, of course. We spend money developing solutions and infrastructure, only to discover that we need to create additional features as requirements change. But – the big difference compared to 1958 is that the modern software development team will expect the environment to change. They prepare the software for integrations, plug-ins, and upgrades. They create roadmaps that pre-empt user behaviour. They allow the system to evolve over time.
So with all these expected changes, how do we know that we’re getting a good return on investment if the goalposts are moving all the time? Can we ever predict what the real cost will be for development, implementation, enablement, and support, of the evolving product?
The case for Agile
Here at DCSL, we are big fans of Agile development. By taking an Agile approach to your software development, you can start measuring the actual ROI and see more tangible benefit to your business, compared to traditional development methods.
- Increased business value. An Agile team can measure the amount of business value in each development sprint, by using measurements for tangible output defined by the project owner. This will ensure each stage is successful before moving on to the next one.
- Increased velocity. The Agile methodology is designed to improve its own efficiency over time. Each lesson learnt becomes a value-add for the next cycle, which continually speeds up the process and gives the client faster value. This can be demonstrated in key performance indicators, showing how the ROI is improving.
- Increased end user happiness. User happiness is a point that many fail to measure, but that is critical for long-term software success. If the user can do more things, faster and easier, they are likely to be happier with the product. This can then reflect in metrics such as productivity, price elasticity – and level of satisfaction.
These are just some of the reasons why we advocate Agile software development. (And we like to think that if the Agile approach had been around in 1958, MOCAS would probably have been a lot easier to modernise today.)
The future of the world’s oldest software
So what does the future hold for MOCAS? Well, the mainframe system is still running, but in 2014 the DCMA issued a request for information to get suggestions for improving and upgrading the entire system. A complete modernisation project is now in progress.