3 ms·
>What we generally mean when we say "nice problem to have" is that the preconditions for having this problem are nice to have. i.e. you cannot have this problem
by coldtea 5y ago
>What we generally mean when we say "nice problem to have" is that the preconditions for having this problem are nice to have. i.e. you cannot have this problem unless some other nice thing happens first. This is not at all a reliable predictor of the amount of suffering involved.
Doesn't matter, because at that point you've already made millions/billions...
That's kind of the whole point: that a "nice problem to have" emerges only when it can no longer be a catastrophic loss of your early startup, but an annoyance in your mature and company (perhaps after you're already sold it, IPOed, and made millions). At that point it's not your problem as a CEO/CTO/etc anymore, it's just a case of throwing money at some engineers to fix the thing.
E.g. who gives a fuck if PHP was not robust of the early architecture not scalable enough for Facebook? They've already made billions by the point it became an issue, and they trivially threw money and engineers and fixed it eventually...
Whereas obsessing over the "right future proof" architecture would have likely killed their early momentum, diverted precious resources, and so on...