Two things are true at once: 1. Apple products work very well with one another 2. Apple products don't work very well (or at all) with non-Apple products I w

Two things are true at once: 1. Apple products work very well with one another 2. Apple products don't work very well (or at all) with non-Apple products I would propose that these are both necessary entailments of the same principle, whose articulation we owe to, of all people, Martin Heidegger, the early-twentieth-century German phenomenologist. In _The Question Concerning Technology_ he called this 'referral', and it's present in all forms of tech, and the degree of referral increases with the complexity of the tech. For example, a fountain pen 'refers' to an inkwell. The more efficiently the fountain pen fits the inkwell -- the less _waste_ there is in the design of the inkwell and the pen -- the _less likely_ it is that another pen can use the same inkwell. (This is my example, not Heidegger's; also, I'm not sure if he noticed that referral increases with complexity -- it's been a couple decades since I read the QCT.) Designs that optimize for interchangeability and multi-part sourcing are _intrinsically wasteful_ -- for example, your Electron app runs slowly because it uses several elaborate engines (Blink, V8) that in turn rely on protocols that are also in some sense wasteful (they aren't binary; they aren't protobufs, etc.). Now, wasteful doesn't mean _bad_, but I propose that in some high-complexity fields where the envelope is being aggressively pushed (e.g. creating new market categories, like 'biometric, account-interlinked, phone-finding, app-running smartwatch'[^1]) it may be a tradeoff that is rejected by the market. Further thinking: Suppose that complexity necessarily increases over time (the Complexity Ratchet -- a persuasive theory I tend to endorse.) If this implies that Referral is always increasing, and does this imply that we are moving asymptotically towards a point where the amount of 'waste' necessary for competition is simply incompatible with the requirements of what is being engineered? This would imply that there is a level of technological complexity after which we lose the substantial benefits of free markets, as competition becomes impossible. I need to think about this more. [^1] Yes, I know Fitbit got to 'smartwatch' first, but I'd argue that the Apple Watch is a tremendous advance over Fitbit, which is why they are preferred.