We talk about scaling a lot, but rarely look past "more traffic" or "ten years from now". This is becoming an issue for the first time because the concept of "

We talk about scaling a lot, but rarely look past "more traffic" or "ten years from now". This is becoming an issue for the first time because the concept of "massively complex computer system at the core of a business" has only been around for half a century or so. Every year it gets more and more expensive (in terms of both people-hours and $$$) to rebuild, refactor, or even improve the system. People who make decisions see the gigantic pricetag on fixing something that already works and opt for patches instead, which is understandable - why fix something that isn't broken (yet), especially when it would cost that much? But as the article describes, when it comes down to the wire and a ground-up refactor is necessary for a system which is 30-60 years old, there may not be enough people who understand the hardware/software, those people are too important to reallocate to the project, or there's simply not enough time/money left to make it happen. I would imagine this new dilemma will start to become much more common in the next 10 years as legacy systems age beyond feasible maintainability.