People who build complex, large-scale, real-world software have an interesting rule of thumb: only about half of the software-building organization’s efforts can be applied to building new features. The other half is devoted to responding to problems of various kinds with the software, an activity that goes by the name of sustaining. Some part of that sustaining work is correcting errors, but other parts fall into a gray zone where the boundary between “bug” and “feature” is trickier to identify.
It’s sobering to realize that there is effectively a large tax applied to software organizations: if you think that something you want to build requires three people to do the actual work of building it, you should hire six people.
Obviously, this rule is helpful for certain analyses of budgets and hiring in software engineering; but beyond that, I find it to be more widely applicable. Embracing this theory and its implications for everyday life can be empowering. It helps me understand an otherwise-frustrating feeling that much of my life seems to be consumed by random things going wrong and needing attention, whether those problems are technical or relational in nature. Indeed, there’s a plausible argument that we collectively underestimate the sustaining work that needs to be done, and wind up cutting corners and creating problems for ourselves at every level from an individual’s life to the stewardship of the planet.
Everything needs sustaining
The underlying issue is that everything of value needs its own kind of sustaining, to either prevent or respond to failures and degradations of various kinds. For example, it’s nice to have friends; but in the absence of sustaining, those friendships fade. For another example, I live in a beautiful house with a beautiful garden – but the dishwasher is starting to misbehave, and one section of the garden is overrun with weeds. My laptop has one USB-C port that doesn’t work; although I have looked up how to troubleshoot the problem, the relevant steps will require me to carve out about fifteen minutes if everything goes well. It will perhaps take hours if things don’t go well, so that I have to take a trip to the Apple store. Even as I write this, I can hear the distinctive sound of a toilet refilling because of a slow leak, somewhere in the house: that situation will require another engagement where I operate as a combination detective and janitor.
I have no basis for complaining about any of these problems: they come attached to my possessions. These problems explain why in practice my possessions own me, rather than vice-versa. The long-term solution to this overcommitment is to reduce the number of possessions to a workable constellation that actually provides value.
Capitalism
Unfortunately, that kind of sensible approach is not the way that capitalism works. Everything that I own has been purchased to solve some problem that seemed important, urgent, or both, at the time I bought it. Few, if any, items came with a warning label for the burden of sustaining I would be taking on. Indeed, there tends to be an opposite set of marketing messages: that something is “maintenance-free” or “lifetime guaranteed.” These claims are not necessarily untrue, but they do tend to obfuscate or omit the very real burden of maintenance that’s associated with the mere possession of pretty much anything.
Maintenance is at best unglamorous, and widely viewed as a negative experience. Items are sold on the basis of being (unrealistically) low-maintenance or zero-maintenance, when arguably it would be healthier to have an upfront understanding of the maintenance burden involved in ownership and how the purchaser will manage that additional work.
Even items that are put on a shelf for potential future use impose a storage and cataloguing overhead, as well as the subtler psychic costs of clutter or the discipline required to avoid clutter. For proof of this claim, we need look no further than Marie Kondo, who has built an entire business around fixing this problem. For myself, I know that I would be well served to think more carefully about what (and who) I bring into my life. Instead of only responding to the many forces urging me to add something (or someone), it’s wise to also consider the inevitable associated costs of sustaining.
Why is maintenance unpopular?
A related question is why maintenance has a bad reputation. I found recently that when I took care of some maintenance task, I felt good. It seemed like an instance of virtue being its own reward. Is this a personal oddity and thus nontransferable? Or have I somehow conditioned myself to this feeling, in which case others could do likewise? Or is it inherent in all successful maintenance activities? In any of these cases, why does maintenance so often seem unappealing?
One negative aspect of maintenance is that it’s inherently repetitive. Even if it’s the first time you’re doing some specific maintenance task, it likely won’t be the last time. In contrast, you may have a genuinely new idea and write it down and publish it, without ever feeling like you’re repeating yourself.
Semi-related to its repetitive nature is that maintenance is often associated with the unoriginal. This association is not really because maintenance should be done by unoriginal people. Rather, this association is probably a consequence of rational allocations of people to tasks. If we assume that everyone can do maintenance, but only a subset of people can be original, then we will assign the able-to-be-original subset to the tasks that require originality. The remainder of the people and the remainder of the tasks will tend to go together: tasks that do not require originality, done by people who are not seen as original.
So what?
What’s the bottom line here? It feels like there are two kinds of personal or societal discipline that can be added or strengthened.
The first kind is relevant at acquisition, and essentially centers on the question of whether the acquisition is a good idea when sustaining is also accurately included in the accounting. This is effectively a kind of “lifetime” or “life cycle” view of the acquisition.
The second kind is relevant for all owned & controlled items, including those which were acquired with no view or knowledge about sustaining. For all those items, the necessary practice is to realistically assess and budget the necessary sustaining. That might well include the need to “de-accession” some possessions if that’s the right choice, recognizing that for anything owned, there is some level of unavoidable cost. Pretending otherwise is just a form of willful blindness that is potentially damaging both to the individual and to the broader society.