About a year ago, I gave in to a temptation and got myself a UPS for my home desktop and the things attached to it, including my DSL modem. When I bought it, I expected it to be almost entirely an indulgence, which I justified to myself as insurance against a very badly timed local power outage while I was in the middle of something important while working from home.

(At work, power outages are too rare to worry about, and one that took down my work desktop would also take down our servers. At home they're more frequent enough to be an actual concern, although still not anywhere near common and would only affect my desktop.)

In practice, my UPS has been surprisingly handy, in two different ways. The obvious way is that it's made the infrequent power outages much less disruptive, and much more than I expected, especially for short power outages of a few minutes. In an ideal world, the impact of a few minutes of power outage on me would just be a few minutes of no computer. In practice, various things make even short power outages much more disruptive than they seem. Not only did I lose any un-saved things and disrupt my ongoing work, but the effects lasted well beyond when the power returned, partly because my DSL modem often took a significant amount of time to re-acquire DSL line signal after it turned on again (and then there could be further delays to bring up the DSL connection). With a UPS, all of that is avoided. Even if I opt to get up to take a break (after saving everything) and wait for the power outage to pass, I can immediately get back to work when power comes back.

The other way the UPS has been handy is that I didn't previously realize how often my area had little momentary power blips, ones bad enough to dim the lights and upset some electronic gadgets. With a UPS, the effect on my computer, my DSL connection, and so on is merely a 'clunk' as the UPS turns itself on briefly. If nothing else, this is reassuring (I don't know if any of the various power supplies were previously being affected by these). I would like to have statistics on how often these occur, but unfortunately they don't show up in the information that the UPS makes available to me.

What I take from this in general is that this is yet again a situation where little irritations and frictions added up to more than I thought, so that removing them had an effect bigger than I expected. And I really should have expected actual power outages to be more disruptive than their mere duration would expect, because everyone knows that interruptions, loss of flow, and so on have outsized effects that go well beyond the minute or two your co-worker will take up. But a belief otherwise springs more or less eternal.