The Plan Was Never the Problem
The most common thing I hear at the Joinery isn’t about a tool. It’s some version of “I don’t know what to build,” and close behind it, “even if I knew, I couldn’t design it.” People say it like it’s the wall between them and the shop. For years I’d answer by pointing them at places to get a plan. I’ve mostly stopped leading with that. The plan was never really the problem.
What stops most people isn’t the absence of a plan. It’s the belief that they need a finished, correct one before they’re allowed to start. They don’t. You need a starting point and the willingness to change it. A napkin sketch counts. A photo and some scribbled dimensions count. I’ve built good furniture off less.
There’s no shortage of starting points, and the first one is other people. At the Joinery that means our Discord — “anybody here built a small bookshelf?” goes up, and in three years I’ve never seen it go unanswered — or bringing the idea in on a Wednesday night so we can stand at the bench and work it out. If you’re not in a shop like ours, the same thing exists online. LumberJocks has been at it for years; post in the design or projects section and someone who’s built the thing usually turns up. Better still, find your local woodworking guild. Most areas have one, they’re easy enough to search out, and woodworkers as a rule are almost aggressively willing to show you how they did something. The point isn’t the platform. Somebody near you has already solved the thing you’re stuck on, and they’ll tell you if you ask.
Then there are published plans. Ana White if you want something simple and clean and beginner-friendly. Fine Woodworking if you want range, from a first project up to things that’ll take you a year. I’ve bought my share of both. I rarely build them the way they’re drawn, though, and it took me a while to be comfortable saying that. A plan is built around its author — their tools, their habits, the way their hands move. It has opinions. Half the time the joinery they chose isn’t the joinery I’d cut, or the sequence assumes a machine I don’t own. So I take the plan as a point of view and bend it toward my shop.
The newest way to get a starting point is to ask software for one, and yes, that means AI. I’ll keep this short because I’ve worn the subject out elsewhere. I’ve handed it a magazine article as a PDF and asked it to turn the project into something I could build: a cut list, a model I could open in Fusion. The article used box joints. I don’t cut box joints on that kind of piece, so I told it rabbets, and it changed them. Five minutes. That’s all it is to me, a faster way to get something on the screen I can argue with.
The furthest I’ve taken it was a table. I built the original out of walnut fifteen, sixteen years ago, no plans, maybe a sketch on a scrap of paper I’ve long since lost. Not a complicated piece, but it had some angles and a couple of curves, and I worked those out the way you do without a drawing, by getting them wrong in the wood a few times until they came out right. The table doesn’t fit our new house. My wife asked me to build it again in cherry.
So I photographed the old one from three or four angles, handed the pictures over, and described what I was looking at: the dimensions, the height, the angles I thought I remembered. I told it I wanted Dominoes for the joinery and a roundover on the top, about the way I’d have told my wife if she’d asked how I meant to put it together. It built a model. The connection dropped more times than I’d like to admit and we went back and forth for a while, but in the end I had a clean model, a measured drawing, and a cut list. If I’m honest, it took about as long as it would have taken me to draw the thing myself. The difference wasn’t speed. The difference was that I could set it running and walk off and do something else while it worked.
Here’s the part I keep turning over, and it may only be true for me. I could do any of that because of the walnut one. I guessed the angles because my hands had already made them once. I could look at the model and tell when a joint was wrong because I’d cut that joint wrong before, in real wood, and paid for it. The software handed me a plan. It couldn’t hand me the thing that let me know whether the plan was any good. I built that the slow way, over a long time, in shops that didn’t have a single one of these tools. I keep telling people the tool doesn’t matter, only the judgment, and I believe that. I’m just less certain than I sound about how a person builds the judgment if they start with the tool already in their hand. That’s my thing to sit with. I’m not handing it to you as a warning.
None of which changes the practical answer, the one I started with. You can get a plan in an evening now, from a person, a magazine, a piece of software, or the back of an envelope. So stop waiting for the right one. Get something down, even badly. Build from it. The wood will tell you where you were wrong, you’ll change the plan, and that was always going to happen anyway.
