In praise of the tool you have already learned

The page-layout software I use is older than some of the people who have asked me why I use it. It crashes in one specific, memorised way. Its handling of footnotes is a standing joke. I have no plans to replace it and I have stopped apologising for that.

What a switch actually costs

The advertised cost of changing tools is the time spent learning the new one, and that part is usually honest — a fortnight, maybe three weeks, and you can do the work.

The cost nobody prices is the thousand small things you knew without knowing. Which menu the thing is under. What the file does when it is nearly out of memory. Which of its warnings matter and which you have safely ignored for nine years. Where it lies to you. That knowledge was not learned; it accumulated, which means it cannot be studied for and cannot be hurried.

I kept a rough log through the last switch I made. Competent in eighteen days. Back to my previous speed in four and a half months. The gap between those two numbers is the whole story, and it is not in any review.

A tool you have used for a decade is not software. It is a set of reflexes that happen to be stored outside your head.

When I do switch

Not for features. A feature list is a description of what a tool can do, not of what you will do, and I have never once been limited by capability. I have been limited by the hours in a week, continuously.

I switch for exactly two reasons. The first is when the tool starts losing work — corruption, silent data loss, a crash that takes an hour with it. That is not a preference, it is a failure, and no familiarity is worth it. The second is when it stops talking to everything else: when the files it makes cannot be opened by the people I need to send them to. Being stranded is a real cost and it compounds.

Being slightly worse than the current best is not on that list. It never has been.

The uncomfortable part

I am certainly wrong somewhere in here. There is probably a tool that would genuinely repay four months, and my rule is good at avoiding bad switches and bad at finding good ones. I accept that trade because the bad switches are frequent and the good ones are rare.

There is also a version of this argument that is just fear wearing a nicer coat. The test I use: can I name the specific thing the new tool does that I actually need, in one sentence, without using the word workflow? If I cannot, I am not evaluating. I am browsing.