The making of software, beer, furniture, and other things.
I'm the creator of Symbology (http://appsto.re/us/9r6Icb.i), BeerRiot (http://beerriot.com/), lots of homebrew, some furniture, and other things. There's more about me at http://beerriot.com/bryan.html
I wanted a nicer handle for the tap serving seltzer from my kegerator, so I grabbed a piece of scrap maple and turned one.
I decided to tackle this video differently. Instead of full instructions on how to make the handle, I described the process mostly at a higher level. To help push me through getting it done, I also limited myself to a clip length of four seconds – find something interesting, say something interesting about it, and move on. This video took probably 20% of the time, or less, to edit. It has some rough edges that could have been polished further, but I think it helped me learn that a lot of the feel of the project work can come through, even when many details are left out.
I’ve been enjoying watching other YouTube builders recently. So, instead of writing up my latest project, I decided to make a video about it.
We left our old laundry rack with friends when we moved. But with cold weather moving in, we have once again needed a place to dry all of our warm wool clothing.
This rack gives lots of hanging space, with plenty of clearance for long garments. It also folds near-flat for easy storage.
Plans can be downloaded for free below. I know that SketchUp is the popular tool these days, but I haven’t learned it yet. The plans are closer to classic architectural drawings, which is the style I’ve used forever.
Let me know if you build this rack yourself, and what you change if you do!
Somehow this “project” never got a write up here. Near the end of September, my wife and I took a spectacular motorcycle ride from California to Wisconsin. Nearly two weeks of spectacular views and interesting places.
With the COVID-19 pandemic surging throughout the US again now, I imagine a few raised eyebrows after reading the previous paragraph. Wasn’t this dangerous, and didn’t we help to spread the disease?
No, we didn’t, and here’s why: we followed the very simple guidelines the CDC and other reputable authorities have been recommending for months.
We wore masks every time we went in a building. We stayed six feet or more from others. We washed our hands often, and carried hand sanitizer for when we couldn’t. We avoided large groups – in fact we even avoided small groups, skipping all the friends we could have visited along the way. We ate outside, most often in empty parks, or alone in our hotel room.
None of our hosts at either end of our journey have developed COVID-19 symptoms. At least one has even confirmed testing negative.
This isn’t hard. Listen to the experts, follow their guidance, and we’ll have this kicked.
I have crossed another long-time TODO project off my list. From the earliest days of learning to play guitar, I knew I wanted to make one. It took me a long time to get started because I wasn’t sure what style I wanted. Instead of cloning of a common model, I decided instead to consider each piece and plan my own design. This quickly led to far more choices and choice-dependencies than I could handle. But this summer, with a little nudge from a friend who also wanted to build a guitar, I took the dive.
Some of my choices had already been made for me. Shortly after I decided to build a guitar years ago, a friend of a friend gave up on his own build. He offered the components he had to anyone who would promise to complete their own build. I’m lucky he didn’t put a time limit on the promise!
These components resolved my choice of pickup, tuning machines, and tail stock. A bridge-position humbucker, 3×3 tuners, and a 12″ radius bar-style tailstock also pointed toward a few common models to mimic. In particular, the Les Paul Jr. caught my eye. My friend also building a guitar just happened to have a regular Les Paul that I could use as reference.
I’ll spare you the step-by-step, since there are already myriad videos of guitar builds to watch. But, I would like to talk about a few of my favorite parts.
To start, all of the wood in this guitar is scraps from previous projects. The neck is leftover maple from my sleigh bed, with inlay made of cherry from my dresser. The body is also cherry from the dresser, combined with mahogany from my coffee table. It was a fun challenge to figure out how to work with what I already had on hand, and makes me feel justified in having carted that wood across the country.
I’m particularly proud of the neck. I chose to make it out of a single, solid piece of maple, instead of gluing a fretboard and headstock onto a neck stick. That took quite a bit of extra planning. I ordered the work on the neck to make each step as easy as possible. For example, I cut the fret slots first, so that the stock was still square, which made lining up the very precise cuts easiest.
I installed the truss rod right after that, from the back, through far more stock that would remain later for the same reason. It was much easier to route the straight, deep slot through stock that would be removed to reveal the headstock than it would have been to route with a headstock sticking up near one end.
And so on. In fact, working on the neck at all was part of the planned order. I wasn’t using a standard neck template, so I didn’t have a template to route the neck pocket into the body. Instead, I finished enough of the neck to define the heel end, then used the usual mortise-and-tenon techniques to mark and chop the pocket out of the body blank before cutting out the shape of the body.
I have mixed feelings about the inlay in the neck and headstock. On one hand, the end-grain of the cherry darkened with the finish, and provides great contrast with the pale maple. On the other hand, one of the shapes I chose was too intricate for my chisels, so it’s a little gappy in its setting. I also cut on the wrong side of my line in the headstock, blowing the entire design. These are both errors that could be fixed by chopping the inlay out and relaying something else. For now, it’s just reminders of the need to practice and pay attention.
The body was a fun experiment. Neither my leftover cherry, nor my leftover mahogany was thick enough to form a guitar body on its own. So, what I did instead was sandwich a mahogany core between two layers of cherry. The cherry also had some nice figuring in it, so I resawed my planks, and bookmatched them. I’m very happy with how the cherry looks, and also with the contrasting strip of mahogany down the middle. From the side, it almost looks like I put edge-banding around the front and back.
Somewhat accidentally, a component of the Les Paul that I copied unintentionally is its weight. This guitar comes in at almost exactly eight pounds. I thought about cutting hollows in the mahogany before gluing up the sandwich to reduce that weight. But the potential complication of needing to remember exactly where those hollows were, when I wasn’t yet sure of exactly where the neck would sit, was a hinderance to just finally getting the project started. Eight pounds is the heaviest guitar I own, but it doesn’t feel too bad on my shoulder.
If you’d like to see some more pictures of the build process, please follow me on Instagram @willthatwork. Instagram is something new for me. I’m still pretty uncomfortable with its features. I was hoping I might be able to use it to find more of an art network, to contrast my mostly tech twitter network. We’ll see if I stick with it. For now, it is also where you will find a demo of this guitar:
I have to give one final shoutout to Pete and Andrew. I’m between shops at the moment, and they offered me time in each of theirs to complete this project. It’s awkward to work in someone else’s shop. Even if the tools are great, learning which are available and how they like to be used takes time. Pete and Andrew were each extremely helpful and tolerant of me adapting to and adapter their workspaces.
We are, we are, we are, we are, we are a quaranteam.
We can, we can, we can, we can avoid Covid 19.
Stay home, stay home, stay home all day, except for nearby walks.
We use Skype and Facetime for all our daily talks.
The traffic has reduced itself to levels rarely seen.
Restaurants serve only takeaway / delivery.
Wear a mask if you must be near any other soul.
Hoarders bought up every single toilet paper roll.
It often feels like the entire world has gone berserk.
When will this be over, when will we return to work?
When will kids go back to school, and grocery shelves be stocked?
If you want clearer leadership, use the ballot box.
Thanks for your assistance in flattening the curve.
Thanks to all essential staff who’ve stuck with it to serve.
We’ve saved those who can’t survive Corona on their own.
Maybe next time we can work on saving the ozone.
By a half-planned chain of events, I’ve spent the last six weeks of COVID-19 Shelter-In-Place over 2000 miles from my woodworking tools. Instead of diving right into a new construction after my dresser, I cleaned and then packed my shop, in preparation for a move. While our belongings have made their way across the country, we have stayed behind to “quaranteam” with a friend-couple, their young son, and their dogs.
We have entertained ourselves with other hobbies: walks to keep everyone moving, cooking delicious meals, reading books, and making music. A few ideas for construction projects have risen during that time, but with few tools and difficulty acquiring wood while maintaining social distance, none of them have been undertaken.
Then one of us saw a post about a geodesic dome made of cardboard. The shape alone immediately captured the attention of the four Xennial-age adults in the house. When we recognized that cardboard was the one material in abundance here, from six weeks of contactless deliveries, wheels set in motion.
Google found a calculator for ordering bits of PVC based on the size and complexity of dome desired. Reverse-engineering that math led to a very simple cut list for a “2V” geodesic dome of paper:
• Ten equilateral triangles, with sides of length A
• Thirty isosceles triangles, with one side of length A and two of length 7/8 * A
Seven-eighths isn’t exactly what the calculator produced, but it’s less than 1% off, it makes measurement simple, and it has worked in my experiments.
The size of the dome that is built is also related to A in a very simple way: the golden ratio. A compressible, bendable material, like paper and tape, worked with common tools like scissors or a box cutter introduced enough error that using many decimal places didn’t make sense. Simplifying to estimating the height at 1.5 * A, and the width (diameter) at 3 * A proved close enough for toy structures.
I’ll include some examples of how specific measurements work out later, but before I annoy people by showing how neatly these work out in Imperial units, I’d like to explain how no particular units are necessary at all. Grab a stick or a string, and I’ll walk you through how to build your own geodesic dome without any arithmetic.
Step 1: Sizing your dome
Figure out where you want to put your dome. Is it a decoration for your desk, or a fort to play in? Find a piece of string, a stick, a strip of paper, etc. that you can cut to the desired width (diameter) of your dome. Before you cut it, find its halfway point, and hold it up in the approximate middle of where you will place your dome. This is about how tall your dome will be (the dome approximates a sphere, so you get one half diameter up from the ground). When you have found a size you like that fits your space, move on to step two.
Step 2: Making your tools
Cut your string, stick, strip of paper to length equal to the dome width that you chose in step one. Then cut that piece into three equal segments. I used a paper strip for my measuring device, so after cutting mine to the full length, I folded it into thirds, and then cut through the folds:
Label one of the cut pieces “8” (eight).
Cut off 1/8th of one of the other pieces. The easiest way to do this is to first find the middle point of that piece. Then find the point halfway between the middle point and one end. Finally, find the point halfway between that point and the end. Cut through that final halfway point. I folded my paper three times and cut through the third fold to do this:
Label the piece you just cut “7” (seven).
Step3: Equilateral Triangles
If the edge of your dome-building material isn’t straight, draw a line on it using a straightedge. Using your “8” piece, mark divisions along your straight edge.
Now for the tricky part. Put one end of your “8” piece right on the left-most mark (or corner) of your straight edge, and angle it up so that the other end is somewhere near where you expect the point of an even triangle would be. Mark a dot on your building material at that end of your “8” piece. Do this a few more times, swinging both a little clockwise and a little counter-clockwise from that spot.
Connect these dots in the arc they form.
Next move the lower end of your “8” piece to the next division mark to the right on your straight edge. Swing the other end up until it crosses the arc you just drew. Mark the point at which it crosses the arc.
Draw a line from each of the division marks you just used to the arc-crossing point you just found. You have just marked your first equilateral triangle!
If you’re building a large dome, and/or working with pieces of material that won’t allow you to get multiple triangles out of one piece, you can skip the next few steps. Cut out this first triangle you have marked, and then use it as a template to trace out nine more identical triangles.
If you’re working with a piece of material that will fit multiple triangles, repeat the arc-crossing process at the right-most division of your straight edge.
Draw a line connecting the points of the two triangles
Using your “8” piece, divide the line between the triangle tips.
Connect the division markers on your straight edge to the division markers on the line between the triangle tips. You have now marked out many more equilateral triangles!
You will need ten of these triangles. If you’ve already marked ten, you’re done. If you need to mark more, try extending your angled lines farther upward. When they cross, they will either make more triangles or diamonds. If they make diamonds, draw a horizontal line connecting the corners to make two triangles.
Cut out your equilateral triangles. Make sure you end up with ten!
Step 4: Isosceles Triangles
The process for the isosceles triangles is the same as it was for the equilateral triangles with one difference: use the “7” piece when finding the arc crossing. Use the “8” piece, as before, to mark divisions along your straight edge, but use “7” to find the crossing point from there.
You will need thirty of these isosceles triangles. Yes, 30.
Step 5: Pentagon Assembly
Time to start assembly. Looking at a finished geodesic dome, the eye is drawn to two (non-triangular) shapes: pentagons and hexagons. I’ve had success with assembling pentagons first, so that’s what I’ll show here.
Collect five isosceles triangles (the ones with two “7” sides and one “8” side). Arrange them in a circle so that all of the “8” sides are pointing out, and all the “7” sides are next to other “7” sides.
Connect four of the “7”-side seams together. A gap should develop in the fifth seam.
Draw the gap together. The pentagon will cup slightly. Seal the seam, and the pentagon will stay cupped.
Repeat this pentagon assembly five more times. You should end up with six pentagons, using all 30 of your isosceles triangles.
Step 6: Connect it All Together
This is where construction will really begin to get unwieldy. If you’re building a large dome, I strongly suggest at least one person to help hold. Two if you can get them.
Collect one pentagon and two equilateral triangles (the ones with three “8” sides). Connect each triangle to two adjacent sides of the pentagon.
From here, connect a pentagon into the space between the two equilateral triangles. This will introduce more cupping, like when you sealed the fifth seam in the pentagon. Continue to alternate pentagons, and equilateral triangles, growing this strip until you have only one pentagon left unconnected (you should have five pentagons attached to five pairs of equilateral triangles). Connect the two equilateral triangles at the end of your strip to the pentagon at the start of your strip. You should have a ring that has a pentagonal hole in one side. Tape the remaining pentagon into this hole, and your geodesic dome will be complete!
Apologies for a lack of build pictures of these steps. The pieces pictured so far are being mailed in an envelope as a small birthday gift. But, here there are laid out ready for final taping.
And here is an annotated diagram of what gets taped where. Purple 1-5 are the pentagon seams. Yellow 1-10 are the remaining alternating-pentagon-triangle seams (9 and 10 appear twice to indicate where the wrap-around connects). Red 1-5 are the roof seams (and 2-5 are duplicated to show where the pieces connect.
And one more shot with a completed dome in the opposite color scheme, to aid in visualization.
If you followed along, I hope your first dome was successful. If you’re wondering about the dimensions of the domes in my pictures, they are these:
Small dome, with blue pentagons: A = 2 inches. Isosceles sides = 1.75 inches. Height is just a bit over 3 inches. Width is just a bit over 6 inches.
Small dome, with green pentagons: A = no idea. I purposely didn’t measure anything, to make sure I wasn’t lying about being able to build this without numbers.
Large dome, made of cardboard: A = 24 inches. Isosceles sides = 21 inches. Height is just a bit over 3 feet. Width is just a bit over 6 feet. The additional ring around the bottom is ten inches tall. We have fit four adults and one child inside. It’s close, but not cramped.
I’ve pushed off writing about progress in the past few weeks, for the practical reason of spending that time in the shop, and for the vain reason of keeping secrets before a big reveal. Last night it became possible to dispense with both reasons at once.
It’s finished! It’s in place. Drawers are filled. The shop can move on to its next project. But before that happens, let’s catch you all up on the intermediate progress.
If you missed the first several steps, these posts will catch you up:
Picking up where I left off, I mounted the drawers on ball-bearing slides. The drawers are just short of 18 inches deep, so I used 16 inch full-extension slides. The small overhang of the drawer above feels natural, like the bit of drawer left inside in a traditional design.
I used Rockler’s slide-mounting jig to ease installation. The only complication I had was matching the flush alignment on the internal dividers to the set-back on the external walls. The top four drawers are essentially inset on the left and overlay on the right. To manage this, I cut a small block to the depth of the inset, and then placed that in the jig ahead of the slide.
With every project, there are intermediate points where things actually look pretty good, and I wonder if maybe I shouldn’t continue with the rest of the plan. At this point, I kind of liked the highlight of the birch next to the cherry. Maybe I didn’t need to continue making faces. A friend even remarked that the blue tape temporary pulls matched well. Continue, I did, though.
The simple face solution would have been to run a plank horizontally across each drawer. I though this would break up the appearance too much, though. It wouldn’t be the simple solution for the cabinet door, either. Instead, I chose to run planks top-to-bottom.
My initial plan had been to glue up this whole panel, and then cut each drawer and cabinet face from it. But, with the assembled dresser taking up space, and the need to keep a path clear for laundry, there just wasn’t room. Instead, I very carefully labeled and cut each piece for each section, and glued each drawer face together individually.
This required some extra attention to alignment along multiple axes while clamping, but I think you’ll see that it all worked out.
I delayed any choice on handles for a long time. Cherry or an accent wood dovetailed into the edge? Leather loops, especially after seeing how the temporary tape pulls fit? I ultimately fell back on my second favorite project material: slate.
Four-inch long, near-square rectangular prisms: one-half inch top and back, Five-eighths inch front, and an approximate ten degree bevel connecting front to back on the underside. I drilled holes an inch to either side of center, into which I gorilla-glued insert nuts.
I used a flagstone sealer to give them a richer tone, and a smoother feel. It’s the same sealer that I used on my coffee table a few years ago, and it has held up well there.
I chose “dark antique” brass butt hinges for the cabinet door. Only a small portion of the hinge is visible, but the dark finish matches the slate well.
Mortising hinges is the second technique I practiced with my box project last spring. The process here was the same: use a marking gauge to layout the cut, cross-chop and clear waste, and lay in the hinge.
To keep the door closed, I embedded a magnet in the door, and a matching one in a small block installed in the case. A one-half inch forstner bit made a perfect hole for neodymium magnets, tacked in place with a dab of Old Brown Glue.
I’ve learned that for eight years, the note about the simple mineral oil and beeswax finish that I used on my bed has had the ratio reversed. If the way is was written, one part oil to four or five parts wax, is correct, then I didn’t pack the wax into the tablespoon at all. I think it’s more likely that the correct ratio is four parts oil to one part wax. That mixup is likely what caused so much trouble finishing those toy blocks last year.
Since I didn’t realize the error until mixing a large batch at the wrong ratio, I had limited options to recover. So, this project’s finish is a two-to-one ratio of oil-to-wax. It took a little more elbow grease to smear on again, but it did smear this time.
In preparation for wax, I first sanded everything to 220. After that, I ran a damp cloth over the wood to raise the grain. When it dried out, I lightly sanded to 320 grit. At this point, I applied a light coat of straight mineral oil. My thinking here was to get the saturation started in the wood, so that fewer coats of oil with wax would be needed. When the oil had soaked in, I lightly sanded to 400 grit. Finally, I applied two coats of oil and wax finish over an eight to twelve hour period, and then buffed off the excess with clean microfiber towels.
I had been a little worried about these dovetails. They’re good, but not perfect. With the wood dry and pale, the gaps were kind of obvious. Oil and wax swelled, darkened, and filled everything. I’m quite happy with them.
We moved it in and transferred my clothes as some final touches were curing. It felt enormous in my garage, and it feels large in comparison to the small dresser it’s replacing. But, I think it does fit the space.
I’m not a great photographer, but I think that some of the curl can be seen catching the sunlight from the window here.
Many of the edges also have a beautiful ray flake that gleams as you pass by.
My sweaters now have a home, instead of piling up on an ottoman nearby. I worked in two small drawers for accessories, including one protected by a lock. These were my chance to include classic techniques as well: horizontal grain orientation (still matched across faces), and wood-on-wood slides.
My first picture in this project’s album, of the wood loaded in a trailer ready to take home, was taken on April 5, 2019. I officially said there was nothing left to do on March 11, 2020. Just over eleven months to complete this project broke down as roughly four hours per week (half of one weekend day) for the first nine months, followed by six hours a day, five days a week for the last two months. That comes out to nearly 400 hours of work. I think it was worth it.
In a decade where it was hard to escape news about how fast technology was advancing, I experienced stability. Wherever I went I carried a laptop. Emacs was always open on that laptop.
This stability led to me finally adopting a journaling system that I could stick with. I just kept a file open in an Emacs buffer, and dropped notes about what I was doing in there as I went about my day. Exact details changed as years past, but mainly:
Plain text, mostly Markdown-ish in syntax.
One file per month.
Start the day by typing day name and date on the next empty line.
Commit the file and push to a remote git repo to backup.
Yes, there are a hundred other tools I could have used. These were the low-energy entry points that meant I kept using it.
That stability has ended, though, and Mac-plus-Emacs is no longer ever-present for me. Now I’m as likely to have only my iPhone or iPad at hand. I still depend too much on digital media and communications to move to a paper notebook, as my father has always carried, no matter how nice Moleskines look.
I’ve tried a new solution for a month now, and I think it’s going to stick: Bear. Other apps came close, but Bear pulled a few important things together in one place: export, sync, price, and usability.
Almost every note-taking application supports note export in some form. Even Apple’s Notes exports to PDF. Bear exports to many formats, and documents doing so in a way to leads me to believe this is a feature they care about. Export to Markdown means that I can maintain my git-repo backup habit for now, and will make it easy to fall back on Emacs if necessary.
Syncing is equally ubiquitous. Bear syncs via iCloud. (CloudKit in particular, which is a bonus point of pride for me.) I love that this means I don’t have to sign up for yet another service, that this service isn’t going to go away if the Bear team does, and I don’t have to think about what data Bear could mine from my notes.
The features I need in Bear are not free. Gaining access to them is $15 per year. But, that price covers Bear on all of my devices. Separate charges for desktop and mobile apps, I understand, helps teams justify the development of each. From my consumer perspective though, I bought access to Bear and now get to use it everywhere, and that’s great.
There are many nice touches in Bear’s usability (two-finger swipe to go to the top or bottom of a note, or automatically pulling a URL from the clipboard when adding a link to a note, to name a couple). But the most important to me is that it doesn’t replace the standard Mac keybindings. My Emacs muscle memory appreciates the basic cursor movement shortcuts supported by all standard Mac text fields.
Bear is working great for me. So much so that I’ve begun to expand usage beyond my journal. The last two blog posts here were written in Bear – the export to WordPress on iOS is smooth. I’m experimenting with archiving recipes, using nested tags.
There are a few places I think Bear has an opportunity to improve, including a bug or two. But they’ve done such a nice job with the fundamentals, and sprinkled just enough extras on top, that it’s already app I’m happy to use. Here’s hoping I get another period of tech stability out of it.
It turns out that there’s a reason everyone recommends building a crosscut sled for your table saw: they really are very nice. I spent the past several days making a few, and it made building my drawer boxes a breeze.
“A few” table saw sleds? Yes. I wanted a full-width sled mostly to fully support some of the larger pieces I would be working with. It also made for a good solution to the problem I have with my table saw: the inset for the throat plate is so shallow that it’s difficult to make a zero-clearance insert for it. I also made two half-table sleds! One for the left side, and the other for the right. The left sled allowed me to use my dado stack without ruining the zero-clearance kerf of my full-width sled. The right-side sled I haven’t used yet, but I imagine it being the small-piece tool after I’m eventually forced to use the left side with the blade leaned over in an acute miter.
Pictured above is the extension I made to cut box joints using the left-side sled. This is another YouTube gem that I now understand why people love. Once tuned, it cuts very even, easy box joints. A testament to this fact is the first picture above, of my top drawer. If you look closely, you’ll notice the corner in the front is labeled C2 on one edge, and C1 on the other. I assembled the side of one drawer with the back of another without noticing – it’s a perfect fit.
I also have to praise the five-cut method. I used it to square my fences, and didn’t appreciate how good it really was until it was proved to me how much more accurate my cuts were that the factory cuts on the ends of these pre-finished baltic ply drawer panels. The box corners that I cut fit like a glove, but the ones where I trusted the factory were off just enough to be annoyingly tight.
The feet were the start of the avalanche. It was time to make the rest of this project three-dimensional. Six pieces would be glued to each other, and two would float in captured grooves.
I had thought through the process at every stage of planning and construction to this point. Then I drew a diagram, and redrew it, and then briefly considered other methods as I started. In the end, the right answer was to borrow an extra pair of hands, so that multiple joints could be managed at once.
That is, referring to the diagram above, 1-3 got me to this point:
The rest of the plan might have worked if I had a large enough space to lay the dresser down on its face. Instead, I had to prepare 4-6 together:
And then fit 1-3 with it:
This is where my choice of glue became important. In previous projects, I’ve stuck with standard PVA wood glue. This has generally been a good choice, but it has a relatively short time between application and setting. With so many joints going together at once, I needed lots of extra time to apply the glue to each surface and then to get all the surfaces into their final position. So for this portion of the project, I switched to liquid hide glue. In addition to a basic long open time, hide glue can also be readjusted by warming it to temperature, so even if I got stuck (sorry!) I would have a second chance.
I warmed the bottle to about 130°F in a water bath, and then applied it to the joint surfaces, using a paint brush to ensure an even spread without too much mess. Fortunately, mess is another benefit of hide glue – in an experiment, it blended into the finish much better than the PVA I would have used otherwise:
Light coat of mineral oil and beeswax applied. Old Brown Glue the clear winner because it vanished in the finish, instead of blocking the oil from the wood like the Titebond. pic.twitter.com/phxn8XR2VT
With my wife’s assistance, we got everything lined up and pushed together. The dovetails slid together nice and tightly with just a bit of clamping pressure. One more round of glue application, and the top followed.
It turned out that the square clamps were not truly necessary. The case stood square on its own. I attached them anyway, in case the glue created uneven pressure while curing.
Except for shaving the overhanging edges of the dovetails flush, the case is now complete. On to drawers!