A problem known as "technical debt" - where poor programming and architectural and coding practices cause problems later in the life cycle of a programme – is putting organisations and businesses at risk. That’s the not-entirely-unexpected finding of a recent study by Cast Software, makers of tools for monitoring software quality.
But in a potentially embarrassing finding for all of those in
Java developer jobs, it was Java apps that had, on average, the most “technical debt” – which is to say, the most flaws that will cost the most to fix in the future. Such flaws can reduce software performance and increase the chance of security breaches, data corruption or system crashes, and the cumulative cost of fixing each line of code in a software application is its technical debt.
Cast looked at applications made by 160 separate companies in 12 different industries, searching for 1,800 types of development violations in applications written in Java EE, Cobol, .Net, C, C++ and other programming languages.
By adding up the number of flaws and then calculating the average monetary cost it was found that Java EE fared worst at US$5.42 per line of code, while Cobol did best at US$1.26 – probably because Cobol has been around significantly longer so many of the most critical errors are well documented.
It’s thought that Java came off worst in the study because of its relative ease of use. Many people in
IT contract jobs code using Java, despite having a relatively weak understanding of the principles of good software engineering.
