On 22 October, the UK government warned us to ensure we had stocks of canned food, bottled water, and other provisions to see us through difficult periods in the coming winter, when supplies and services might be disrupted. For a nation that survives on coffee and tea but is woefully short of camping stoves, that could prove a real challenge, and if you’ve just traded in your old car for an EV you might now be having second thoughts.
As we learned on Friday, it’s not families and other folk who lack resilience, but government, businesses and other organisations. When Microsoft’s servers became unavailable, hospitals and healthcare services, courts, transport systems, and enterprises around the world were unable to function. I hear similar cries from developers when their access to Claude or other AI is unavailable. Over the last few years, many have been handing their daily business over to remote services and their servers, paying dearly to put all their eggs in a single basket, and throwing resilience away.
Although a far cry from Microsoft, AI and other providers, Apple isn’t exempt. It still promotes putting your Desktop and Documents folders into iCloud while enabling Optimise Mac Storage, and renting terabytes of remote storage, so increasing your exposure to service disruption. But at least those are options the wary can manage or avoid, rather than being requirements of Mac ownership.
Resilience is dividing your eggs between several different baskets, making those as independent from one another as much as possible. It’s having not just plan A, but B and, if possible, C as well.
Mains power is a common requirement, although it’s amazing how much you can still get done on a MacBook Pro while you’re waiting for the lights to come back on. As a former ham radio operator who enjoyed operating mobile, I used to keep a small generator handy, but that’s only worth the cost and effort if you’re out in the sticks where outages are relatively frequent.
Planning to work for a few days without a high-capacity internet connection should be more feasible. That means you can’t afford to be dependent on off-device AI, and use it as an enhancement rather than making it central to your work. I have a great advantage here, in that my commercial writing for MacFormat and MacLife explicitly forbids the use of any form of AI to create my work. I just wish that were true of others who publish what they haven’t even checked for themselves.
A fallback Mac is almost essential unless you could get by temporarily with your iPad or iPhone. I do perform some administration of this blog using the Jetpack app on my iPhone, when I’m out walking on the local Downs, for example. That’s sufficient for providing brief answers to questions that don’t require any research or checking, and for trashing spam comments. But even lightweight editorial work such as correcting typos is better left to my mainstay editor, MarsEdit, which I have used to write and maintain every single one of the 11,196 articles now in this blog.
Although I’d hate to disappoint you and not publish when you expect, meeting deadlines is far more important for magazines. To ensure I have good fallbacks, I upload my drafts to iCloud Drive as I work on them, as well as keeping hourly local backups and more frequent saved versions. A week before my deadline I assemble all the material I need with an outline of the contents, and send that outline to my editor for their approval. From then on, it’s just a matter of transforming that into structured text ready for production. As I work throughout in plain text, using BBEdit, I’m the only potential obstacle to meeting the deadline.
Resilience is all about minimising the risk of failure by building several independent paths to get things done. In basic probability, the combined risk of two or more unlikely events both occurring at the same time is the product of their individual risks. By adding more separate and independent paths to your workflow, you keep reducing the risk of failure. With a choice of three plans, you can only fail when all three fail together.
What we should learn from Microsoft’s service failure is that placing complete reliance on a single plan that has multiple and mutually dependent risks built into it is doomed to fail sooner or later, and more often. It’s putting all your eggs in someone else’s basket, then wondering why they get broken.
Perhaps that’s why the UK government is warning us to prepare for disruption of services, because those are no longer resilient as they now rely on remote services such as Microsoft’s, and have no plan B.
