It wasn’t the MacBook Air’s SSD problem confounded by encryption usability problems and the “Update Needed” ghost user. 
It wasn’t the cognitive gymnastics that connected inability to access iCloud video to Apple’s newly announced device limits. (Though once I connected this to weeks lost to iTunes sync bugs I was probably getting there.)
It wasn’t that our still warranteed AirPort Extreme Base Station acts like it has a failing power supply.
It was, finally, when the MacBook Air ran into a cyclic reboot problem. That’s when I did the math.
Our family owns 5 iPhones  and 3 Macs (and various iPods, but I’ll ignore those). Child #1 and #3 have school iPads. About ten devices across five users, and each user has iCloud and Google Accounts (more than 13 Chrome Profiles). So maybe 20 or so things each of which has a 98% chance of being problem free in any particular week. How often should we have a trouble-free week?
That would be (0.98)^20, or about 67% of the time. So about 1 week in three I should run into one or more significant debugging problems. That’s pretty much what I see. I have other things I’d rather be doing.
This isn’t the first time I’ve run into this kind of complexity crunch. Until 2005 we were primarily a Windows XP household with a single lonely iBook. XP was emphatically not problem free 98% of the time. Maintenance was eating way too much of my life. I bought a G5 iMac, retired the XP machines, and, after I made it through some grim early days with the G5, life got a lot better.
So what can I do in 2015? XP was pretty bad by 2005 — I don’t have such an easy target today. I’ve already cut out a lot of services; we use a selective mix of Google Apps and iCloud with a handful of other high quality high reliability services (Pinboard, Feedbin) that only I use.
The answer, I think, is fewer devices. So instead of buying an iPhone and an iPad, buy an iPhone 6s+. If I want a new MacBook, I have to find something comparable to get rid of. If the WiFi is bad in a part of the house, I don’t buy a WiFi extender; I just don’t use the WiFi there. Over time, work towards fewer devices and services — sacrifice power for reliability.
Oh, yeah, and no (useless) Apple Watch. Life is too short.
Fewer devices means it’s time to modify Gordon’s Laws of Acquisition (2008)…
- Never acquire anything until you really, really, want it -- three separate times.
- The real cost is the lifetime cost, from acquisition to disposal … think subscription — not ownership. In the modern world we don't own, we subscribe to something that's neither inert nor living. The purchase price is often the least of things.
- Don't buy on promises or potential. Acquire for real value now. Anything in the future is a plus (or, sometimes, a minus).
- Don't buy more than you can consume now. We all have fixed resources to acquire and adopt new things; acquisitions that sit on the shelf depreciate very quickly.
- (new) Every purchase must reduce maintenance time and complexity, typically by replacing a less reliable device or by substituting one device for two devices.
- Why you will live in an iOS world 12/2010: “a hunch this opportunity cost is important to understanding what happened to the world economy between 1994 and 2010” foreshadowed a post from two weeks ago. Contrary to what i wrote then I think I’ll still be on OS X in 2018.
- Gordon’s Laws for software and service use 4/2010. i don’t pick up new services any more, not unless I can get rid of 1 or 2 old ones.
 I won’t pay for anything else. The thought of trying to maintain any other type of phone gives me hives.
 My blog post is still in draft. That was just 3 weeks ago.