I think technical debt extends beyond poor quality code that prevents additions or modifications. Having worked with a large amount of scientific software it ex

I think technical debt extends beyond poor quality code that prevents additions or modifications. Having worked with a large amount of scientific software it extends into the efficiency and robustness of software. Frequently, efficient methods or parallelization are not used and the least creative method is implemented to solve a scientific question. Functions are not written in a modular way but merely for a “quick and dirty” analysis. The end result is that you spend more time getting things to run or waiting for them to run than if you had gone back and reworked the code to begin with. It’s a serious drag on productivity.