Technical Debt

Technical Debt is unfortunately a terribly abused metaphor. Somewhat ironically it was coined to use business/accounting concepts to make technical considerations more accessible but it is those concepts that seem to be either ignored or misunderstood such that Technical Debt is often treated as “anything technical that we wish didn’t exist” rather than things which should be able to be treated as debt.

I also disagree with some of the use of the term in regards to shipping things such as minimum viable products or architectures as it is used here{atime-“2024-09-30”}. While much of the advice in the article is sound, it seems like further dilution of the metaphor: not building what you think you may need in the future but don’t right now is not debt, that’s being empirical and agile. Metaphorically this is analogous to making a smaller purchase than may be desired (but may not be needed) but there’s no indication that it isn’t paid for and therefore would incur a liablity. While the idea that it is something that needs to be managed and may not need repayment (particularly given the prospect of what amounts to feature bankruptcy) is all valid, it is important not to conflate the state of solutions that expected to be “good enough” for the time being from those that are not but for which those time to market takes precenence over some clear non-blocking issues.

Years ago I had a more lax perspective and bought into the quadrant model mentioned by Fowler, but I then witnessed a culture where that led to mismanagement of such debt; without differentiating accrual of debt from any undesirable expenses it can be easy to stumble your way into insolvency which in technical terms can translate to thoughts such as “it doesn’t matter if we cut all of these corners for this new system since we’ll just add that to the backlog along with sunsetting our legacy systems”.

There’s also a somewhat related concept of civilization debt. While this certainly has different underlying motivations, those motivations appeal to me and I’d also aver (without particular support) that many of the debt management strategies would overlap.