I was standing in my Minneapolis garage on February 10, 2026, staring at a 1/8-inch gap in a mortise and tenon joint that was supposed to be the structural foundation of a new cherry coffee table. In my day job as an IT project manager, we call this technical debt. You ignore a small logic error early in the sprint, and by the time you reach deployment, the whole system crashes. In woodworking, the 'crash' usually involves a lot of wood glue, a few choice words, and the realization that you just spent $187.00 on lumber that currently looks like expensive firewood.
I started this cherry coffee table project on January 15, 2026, after visiting my local hardwood dealer. I picked up 22 board feet of cherry at $8.50 per board foot. It was beautiful stuff—creamy sapwood, deep reddish heartwood, and that distinct smell that makes you forget you’re working in a freezing suburban garage. But beauty doesn't compensate for a lack of precision. By the time I reached the assembly phase in late March, I had accumulated a series of 'bugs' in my joinery that required some serious debugging.
As someone who spent years building a lopsided bookshelf before finding my footing, I’ve learned that the difference between a 'failed' project and a 'character-filled' one is often just a clever patch. Here are five joinery mistakes I made on the cherry table and how I managed to debug them before the final finish went on April 5.
1. The 'Off-by-One' Tenon (The Loose Fit)
In programming, an off-by-one error is a classic. In the shop, it’s when you get a little too aggressive with your shoulder plane and suddenly your tenon slides into the mortise like a hand into an oversized glove. A good joint should have a 'piston fit'—it should slide in with firm hand pressure and hold its own weight. This one? It rattled.
The Debug: I treated this like a shim. I took a shaving of the same cherry scrap from my bench—literally a paper-thin translucent curl from my hand plane—and glued it to the face of the tenon. Once the glue dried, I carefully pared it back down until the fit was snug. If you use the same wood species, the 'patch' is virtually invisible because the grain orientation matches. It’s essentially a hardware hotfix that restores structural integrity without having to recut the entire rail.
2. The 'Syntax Error' Gap (The 1/8-Inch Void)
On March 22, I was dry-fitting the base assembly when I noticed a glaring gap where the apron met the leg. It was exactly a 1/8-inch gap, likely caused by a slight twist in the leg that I hadn't accounted for during the initial milling. In the IT world, this is a syntax error—the components are there, but they aren't speaking the same language.
The Debug: For gaps up to a maximum gap width of 1/8 inch, you can sometimes use the 'sawdust and glue' trick, but for a high-end wood like cherry, that often looks like a dark, muddy scar. Instead, I used a 'Dutchman' or a thin wedge of the same wood. I cut a tiny sliver of cherry, tapered it slightly, and tapped it into the gap with a bit of glue. After it dried, I flushed it up with a chisel. Because it’s actual wood grain filling the void rather than a paste of dust, it takes finish much more naturally. It’s a bit like refactoring code—it takes longer, but the end result is much cleaner than a quick-and-dirty patch.
3. Out-of-Square Glue-Ups (The Systemic Failure)
There is nothing quite as soul-crushing as applying glue to a complex assembly, tightening the clamps, and realizing the entire structure is skewed by a quarter inch over its length. This usually happens because the clamps aren't parallel or the floor of my garage—which is definitely not level—is throwing off my perspective. I learned early on that investing in decent clamps is vital, but even the best tools can't fix a lack of geometry.
The Debug: If the glue is still wet, you can use a 'diagonal clamp' to pull the assembly back into square. I measure the diagonals of the table base; if one is longer, I place a clamp across that long diagonal and apply pressure until the measurements match. If the glue has already set, you’re looking at a structural redesign. On a previous project, I actually had to plane the bottoms of the legs to different heights just to make the table sit flat on the floor—a shameful 'hard-coded' fix that I hope no one ever notices.
4. The 'Buffer Overflow' (Too Much Clamp Pressure)
I have a tendency to over-tighten. I think it’s the engineer in me—if tight is good, tighter must be better. But too much pressure can actually starve a joint of glue or, worse, bow the wood. On the cherry table top, I clamped the boards so hard that the center of the panel began to lift off the cauls. I was essentially creating internal stress that would eventually lead to a crack.
The Debug: The fix here is preventative maintenance. Use 'cauls'—straight pieces of scrap wood wrapped in packing tape—placed across the width of the panel to keep it flat while the pipe clamps do the heavy lifting. If you’ve already bowed the panel, the only real fix is to rip the glue line on the table saw and try again. It feels like a rollback to a previous version, but it’s better than a tabletop that looks like a Pringles chip. According to the Forest Products Laboratory, wood is constantly moving due to humidity, so any internal stress you 'build in' with clamps will only be magnified when the Minnesota seasons change.
5. The 'End-of-Life' Snipe (The Planer Scars)
Snipe is that annoying gouge at the beginning or end of a board caused by the planer’s rollers. It’s like a legacy bug that you forget about until the very end of the project. I had some nasty snipe on the undersides of the cherry aprons. I thought I could sand it out, but cherry is notorious for showing every scratch and dip once a finish is applied.
The Debug: If the board is already cut to length, you can't just cut the snipe off. My solution was to use a hand plane to 'blend' the snipe into the rest of the board. It requires a very light touch and constant checking with a straightedge. In the future, I’ve learned to feed a 'sacrificial' piece of scrap wood through the planer immediately before and after my actual project pieces. This tricks the machine into 'sniping' the scrap instead of my $8.50-per-foot cherry. It’s the woodworking equivalent of a sandbox environment—test the dangerous stuff on non-production data first.
Lessons from the Workbench
By April 10, the table was finally finished. It’s not perfect. If you look closely at the bottom of the left rear leg, you can see where I shimmed a tenon. If you crawl under the table with a flashlight, you might find a trace of that 1/8-inch Dutchman patch. But to my wife and our guests, it looks like a solid, professional piece of furniture.
- Woodworking is 10% cutting and 90% problem-solving. No project goes perfectly. The skill isn't in being perfect; it's in knowing how to hide the evidence.
- Cherry is a forgiving teacher but a harsh critic. It machines beautifully, but it burns easily if your saw blade is dull, and it shows every glue fingerprint.
- Documentation matters. I keep a notebook in the garage where I record my 'bugs' so I don't repeat them. It’s my own personal version of a ReadMe file.
Woodworking in a suburban garage involves a lot of trial and error, especially when you're self-taught via late-night video marathons. But every time I fix a gap or shim a joint, I’m building a mental library of solutions. If you’re just starting out, don’t let a mistake stall your project. Most 'bugs' in the shop can be patched with a little patience, some matching scrap wood, and the willingness to admit that your first attempt wasn't quite 'production-ready.' For more on how I've navigated these learning curves, check out my experience with the 12-weekend dining table build, which was its own masterclass in error handling.