Hacker News Reader

> you are unlikely to correctly predict in what way your solution will be inadequate

but you can choose a current solution that does not block nor inhibit a future solution. As they say, always keep your options open, and don't vendor lock-in yourself into a corner.

> Your solution will prove inadequate in the future, but you are unlikely to correctly predict in what way your solution will be inadequate.

One inadequacy could be complete dependency on a vendor who accidentally locks you out of all your services or starts scalping you with ever higher fees. That one is quite predictable here.

But: if it is not more work, you can solve your problem by solving a family of problems around your problem in the problem space, making the solution robust, as small changes in the problem definition can easily be catached by small changes in your solution.

That should, with no extra work, solve future problems.

I think you are saying you should be structured well enough to change your system to solve future problems.

Like do you really need a dashboard on day 1? Or reports on day 2? Sometimes you just need a log of data. And the columns can grow and shrink.