Four hours of print time spent on a part I could have cut in twenty minutes
Choosing a process out of convenience rather than fit, and what it cost.
Printing a part is not always the fastest way to make it, and a flat bracket built up layer by layer on a printer can take hours precisely because the machine has to repeat itself hundreds of times over, once for every layer of height, regardless of how little material each layer actually contains.
What is actually happening
Printing builds a part one thin layer at a time, and the machine has to finish an entire pass across the shape of that layer, however small the shape is, before it can start the next one, so the total time a print takes depends far more on how many layers tall the part is than on how much plastic it actually contains. A flat bracket only a few millimetres thick can still take hours if a large part of that height is spent on a fine surface one whisker-thin layer at a time, because the printer is repeating its slowest, most careful pass over and over rather than removing material in one continuous motion the way a cutting tool would. None of that repetition is wasted effort in the sense of doing anything wrong, the printer is doing exactly what it was asked to do, precisely and patiently, it is simply that patience and precision are the wrong currency to spend on a shape that never needed them in the first place.
The fretsaw comparison
Cutting the same flat shape out of a sheet of plastic with a hand fretsaw takes roughly the same amount of time whether the sheet is thin or twice as thick, because the saw is following the outline once, in one pass, and the blade cuts through the material's thickness as it goes rather than building that thickness up stroke by stroke. A printer asked to make the same flat shape has no equivalent shortcut, since every additional fraction of a millimetre of height means another full lap of the nozzle around the same outline, and a shape a saw clears in one pass can cost a printer dozens or hundreds of those laps stacked on top of each other. The comparison is not about which tool is better in general, since the saw can do nothing at all with a shape that needs an internal cavity or a curve reaching back on itself, it is only that a flat outline, the one job a saw is built to do supremely well, gains nothing from being handed to a machine that solves a completely different kind of problem.
The number that matters here
A single layer on a well-tuned printer might take anywhere from under a minute to several minutes depending on its size, and a part with even a modest few hundred layers of height can therefore take hours to finish something a saw, working through the material's thickness directly rather than one thin slice at a time, would clear in well under half an hour.
What this changes in practice
Choosing between a printer and a hand tool is really a question of which one matches the shape being made, since a printer earns its hours on a shape a saw could never cut at all, an enclosed cavity or an overhanging curve, while a shape a saw or a blade could remove cleanly in one pass gains nothing from being built up layer by layer except a much longer wait for the same result. Reaching for the printer by default, simply because it happened to be the tool already switched on, is a habit worth noticing before it costs an evening rather than after.
My key error with this
The bracket was a flat plate about five millimetres thick with four mounting holes in it, and because I had drawn it in CAD the next step felt automatic, so I sent it away to be printed and paid for the privilege of waiting most of a day for it to arrive. The geometry had no curvature anywhere, no internal features, nothing that could not have been produced by marking out two pieces of sheet and cutting them, which would have taken twenty minutes with tools that were already in the room. What had happened is that modelling the part in CAD quietly selected the process for me, since a finished model feels like it is asking to be printed, and I never went back to ask what the part actually needed. What replaced the belief is that the tool I happen to have designed something in has no authority over how it should be made, and that a part with no features a saw cannot produce is a part a saw should probably produce.