About overthewireless

One time historian turned construction lawyer, musician, martial artist, photographer, distance runner, builder of musical instruments. Hand tool user all the time, every time.

The Final Piece of the Puzzle


Holdfasts have been central to the way I have worked over the past four years, and they are a critical workholding strategy for the Roubo bench. A hefty bench, such as my oak Roubo build, deserves a big hefty holdfast. The distintive holdfasts detailed in With All the Precision Possible are much larger than many modern designs, and I wanted something which would fit the overbuilt vibe of the bench. Specifically, I wanted a Crucible Tool holdfast, or one of Peter Ross’ amazing full scale “Roubo” holdfasts.

Unfortunately Crucible do not ship internationally, but a tip-off over Instagram led me to Hyvlar – a Swedish tool seller who stock Crucible tools and are happy to ship across Europe. My Crucible holdfast arrived today, and while I do not yet have any 1″ diameter holes to test it in, it appears to be an impressive tool. The design is very similar to the engravings in Roubo’s manuscript, and the large size and rough cast texture will complement the oak bench nicely. Of course, none of that matters unless the holdfast can steady workpieces, but everything I have heard from folk who use these holdfasts is that they hold like the dickens. I’m looking forward to putting this holdfast through its paces, and will report back once the Roubo bench is in use.

I’ll be keeping my current 3/4″ holdfast, and may well bore a 3/4″ diameter hole in each of my saw benches to assist with holding stock in place.

An unexpected chest


Last weekend we took a trip to North Shields to see some very close friends. Travel has many benefits, but as I’ve written about previously, staying in holiday cottages often provides the opportunity to get up close and personal with vernacular furniture. This trip was no exception, and I was delighted to find in the lounge of our delightful flat (which was only meters from the beach) an old dovetailed chest. Now, I can’t resist the opportunity to spend some time looking at utility dovetails and this chest was intriguing. It was a touch smaller than my Anarchist’s Tool Chest, but both chests share design DNA. And while the execution was one the utilitarian side, it was clearly built by someone who understood furniture making and woodmovement in particular.


So what caught my eye on this chest? The dovetails are an obvious space to start. They were irregularly spaced, and while some were very gappy, there were also a number of splits radiating from the corners of the chest which suggest that some of the joinery was over tight. Nonetheless, the chest felt very solid, which goes to show that a dovetail does not have to win any beauty contests to be stout. The dovetails were almost certainly cut by hand, as the baseline was still clearly visible on several corners. Whoever built the chest had also used beading to protect the fragile edge below the lid. Tearout and tool marks were also present on the external and internal surfaces.



The lid was comprised of two pieces, which had been joined with tongue and grooves. Opening the lid revealed two rectangular patches on the inside face, where the wood had retained its original colour. Those patches (which crossed the two boards) were pierced at each end with a hole, and this suggests that battens had been screwed or nailed across the lid to prevent warping. The end of one dovetail in the carcase also showed that the sides of the chest had been built out of tongue and grooved boards.


The corners had also been reinforced with metal plating, although some edges of the plates had been folded over, so I am not sure how robust the reinforcement was. Interestingly, some effort had been taken to clock the screws when installing the hardware.



The chest was full of records and casset tapes, and I was interested to find a fixed till at oneend which appeared to be original, and which contained two drawers.


All in all, this was a lovely example of a “user” piece of furniture, and one which has plenty of life left in it.

Interative Design

The following is based on an article originally published in issue 284 of Furniture & Cabinetmaking.

In 2017 a client commissioned me to build an open cupboard for the hallway of a Victorian house in London, with the key design criteria that it was to be able to provide ample shoe storage. The result was the Policeman’s Boot Bench – a dovetailed oak cabinet with four shelves for even the most enthusiastic shoe collector. No sooner had I completed the commission and handed it over to the client, then Dr Moss pointed to the overflowing shoe rack in our own hall and requested that I build a variation on the piece I had just delivered, to address our shoe storage needs (the finished piece can be found here).


Iterative Design

The process of designing, and building, a second interpretation of an existing project provides an excellent opportunity to review your design process: both in terms of how small design changes can have significant impact on functionality, and how design choices affect the overall form of the piece. Subsequent approaches do not rewrite the base design, but adapt and mould it to new settings and uses that reflect the specific needs of the end user.


Despite my best efforts, I have been unable to find a design term that reflects the process of adapting a base design for different functionality or aesthetic criteria, and so for the purposes of this article I will refer to it as “iterative design”. If any reader is aware of a design term which better reflects this methodology then please drop me a line and let me know!

Compare and Contrast

While I will be using the two shoe cupboards to illustrate the benefits of iterative design, really the same process could be applied to any furniture form. In fact, on my workbench are two child-sized Welsh stick chairs. These chairs are both based on an 18th century example, but thanks to changing some design elements (one has a round leg profile while the other uses octagonal legs) and material choices (one is oak, the other maple), they have radically different appearances. Returning to the shoe cupboards:

Version One (the original Policeman’s Boot Bench) is all oak construction, with four full-width shelves housed in dados. The finish is blonde shellac and black wax.

Version Two (which resides in our hall) has the same overall proportions, save for being slightly narrower (11.5” compared to 13”). Pine is used throughout, and the joinery (dovetails, dados, and tongue and groove for the backboards) is exactly the same as for Version One. However, there are three shelves in this iteration of which only the bottom shelf is full-width. The other two shelves terminate in a vertical divider 12” from the righthand end, resulting in a tall compartment at that end. At the top of the compartment is a dovetailed drawer for keys, post, and the inevitable clutter that seems to accumulate by the front door. The shelf spacing is also different compared to Version One. In terms of finish, the exterior of Version 2 is painted with Bayberry Green milk paint by Old Fashioned Milk Paint, topped with Osmo PolyX, while the interior is also treated with Osmo. The dados are pinned with “Roman” style nails from Rivierre Nail Factory, while a Mackintosh brass pull from Horton Brasses allows the drawer to be accessed.


Cutting the dovetails

So those are the differences, but what design choices led to them? This can be split into two elements – functionality and aesthetics.

Functionality – adjusting the functionality of the shoe cupboard was where negotiations started for version two. Our existing shoe rack (by a popular Swedish home furnishing company) struggled to accommodate anything larger than a standard work shoe, and was downright hopeless at holding the Apprentice’s tiny shoes. So, from the very start we knew we needed shelf spacing that would work for my Dr Marten boots as well as a three year old’s favourite sequinned wonders, and Dr Moss’ knee-high boots. The addition of the vertical partition provided the tall boot storage without creating too much unusable dead space, and left me free to have shelves in the main compartment. Having determined the space required for the Dr Marten’s (8 ½” in case you were wondering), the shelf spacing for the main compartment was stepped off with dividers to achieve a gradually reducing space for each shelf up, without the need to deal with awkward fractions. A small drawer at the top of the full height compartment added an extra level of functionality as well as some visual balance to the design – before adding the drawer, the right-hand compartment looked cavernous and disrupted the proportions of the piece.


Aesthetics – the Policeman’s Boot Bench was designed for a Victorian property with an original tiled hallway, and the emphasis on oak and a simple finish fitted that setting very nicely. Thanks to a large lime tree in the front garden our hallway has a tendency to be gloomy, and in an effort to mitigate this we have painted the walls a bright yellow. For this reason we wanted to avoid adding large pieces of furniture in dark wood, and so pine was chosen for gentle tones it would add. The overall form of the shoe cupboard is understated, and relies on hand cut joinery as well as an elegant foot detail to provide visual interest. Milk paint is the perfect finish for adding colour while keeping joinery visible, and the benefit of a mid-toned colour is that the natural wood of the interior illuminates the shelves, making the overall piece “pop”.

These are all simple choices – asking what we want to use a piece for, and how it will sit in the intended setting. And yet, when added together, they result in substantially different pieces.



Cleaning out dados

The Benefits of Iterative Design

Having found the iterative design process to be beneficial and thought-provoking, I was interested to see if other makers used this process. Canadian furnituremaker Rich Wile makes extensive use of this approach, particularly when deploying specific joinery solutions. Rich explained that while he developed a “chair base design using bent laminated braces” for a dining chair, “it has provided the design and build basis for many other projects including benches, stools and small tables. Each application requires small tweaks to the design, such as the size and angle of the brace, but the technique remains the same”. Carver Danielle Rose Byrd also makes use of iterative design, particularly with spoons and shrink pots which “are limited to more nuanced changes in design that benefit from an iterative design process”.

I have found two main benefits to this process, although I am sure there are plenty of others I’ve not yet stumbled upon. Firstly, once the base design has been prepared, creating further iterations is a more time-efficient process, because the overall form (proportions and joinery, not to mention the construction process) have already been established. Instead, subsequent design time is focused on making adjustments to achieve a different functionality aesthetic, or to accommodate different materials. Those efficiencies are not confined to the time spent on design either. As Rich explains: “I tend to over-engineer a piece with extra structure or support the first time around. But the experience of each subsequent iteration allows for refinement or efficiencies to develop in the built to either save material or time, or to increase the flexibility of the design”.


Dovetails for the first corner

Secondly, having an established based design already built (iteration one – in my case the Policeman’s Boot Bench) enables the impact of each design change to be readily compared and contrasted. Again, that might be in terms of functionality, or appearance. By expanding the uses of a base design, you develop an understanding of what functionality can be achieved within a given footprint or internal fitout. This reflects Danielle’s experience, as “it’s one thing to see a new pattern carved onto a flat board, but being able to use an actual shrink pot so I can see how the pattern relates to the shape of it allows me to make better informed decisions about what ultimately gets used on the final piece”, and for this reason she keeps all of her “practice boards and failed past projects”.

The Drawbacks of Iterative Design

No design process is perfect, so are there drawbacks to iterative design? The obvious one, to my mind, is that any process based on repetition or making adjustments to a base design, has the potential to limit design language and result in a narrower breadth of work. Danielle’s experience would support this, and suggests that “iterative design holds me back – I tend to think smaller, be more conservative, and limit myself”. Richard adds that “I think you can become too comfortable with your ‘hammer’ and perhaps have a blind spot for other approaches”, although staying interested in other makers’ designs and approaches can counterbalance that risk.



Within the context of a broader approach to design, I’ve found that iterative design can be a useful tool, and one which has some very beneficial attributes. What I find really interesting about using the Policeman’s Boot Bench as a base design is that making changes to the internal layout requires only moving the location of the dados, which really emphasises the flexibility of both the joinery and the design. As Rich observed about his own experience with iterative design “when the project I am building can [make use of an established design feature], I know I can rely on solid experience and approach, and focus on details and aesthetics and not on behind the scenes building techniques”.

At its root, this iterative design process is about developing a design language and experimenting. In fact, the more I dig into this as a process the more I find that I am engaging in a design conversation with myself and the overall brief of the project. Building that design language presents the opportunity to explore ideas, and draw upon previous experience to trial new approaches As Danielle explains, “when I first got into bowl carving, I set aside ten or so maple bowl blanks and let myself go wild on them, entertaining any and every idea I had so I could find the extremes of what was possible, including texture. I also have bowl failures that I keep around. This comes in handy during the iterative design process because it allows me to easily expand on previous ideas that I’ve hashed out on these testers seemingly picking up where I’ve left off”.


Bowl by Danielle Rose Byrd

This represents my first forays into iterative design, and the experiences of some other makers. And so, I shall finish with a challenge – next time you come to design a piece, ask yourself if you need to start from scratch, or whether you have already built a base design that could be adapted. This approach may save you time and provide some unexpected results.

In Print the “hard way”


Issue 285 of Furniture & Cabinetmaking is now in print, and mark the launch of a 12 month series I am writing about building the slab top oak Roubo bench (the magazine cover dryly refers to buuilding “a traditional Roubo bench, the hard way”). Every month we’ll be looking at one aspect of the bench build and covering it in detail. If traditional workbenches aren’t your cup of tea, not to fear as Mark Harrell continues his series on saw sharpening, Richard Wile examines a range of pencils for workshop use, and Derek Jones continues his bog oak box build. There is also a fascinating profile on the Gandolfi camera company.


The virtues of the 60 second sketch

Design as a process is something I find simultaneously fascinating and daunting. Much of the apprehension about sitting down to design something comes from the fact that I am terrible at drawing. I mean, most of my stick figures look like they need urgent spinal surgery. One of the most challenging aspects of the rigorous training at Totnes was drawing the outline of our guitars. Those three days spent with a sheet of tracing paper and a pencil, chasing smoothly flowing curves was at times hard going (no templates, no french curves, just freehand drawing and lots of rubbing out). It’s funny, because actually once I move beyond the how do I want this to look stage and into the how do I build it? I find the design process much more enjoyable and less painful, because at that stage it is down to engineering a solution.


The guitar I built at Totnes came out pretty nicely, as did the few furniture pieces I’ve designed from scratch, and to progress as a maker the design process is something I want to feel much more comfortable with. So I’ve been thinking about how to make the act of designing pieces more natural. Mike Pekovich’s excellent book The Why & How of Woodworking makes the excellent suggestion of drawing thumbnail sketches to rough out an idea for a piece before drilling down into measured drawings, and this is an approach Chris has also touched on previously. The idea of a 60 second sketch appealed as an efficient way to capture the essence of an idea. Such a quick sketch would be rough by nature, and so mitigates against my lack of artistic prowess.

So, on a whim I picked up a sketch book from our local newsagents. Nothing fancy, just an A5 sized fliptop pad with 5mm squared paper. Not long after, Dr Moss and I were planning the next round of decorating (we are slowly but surely eliminating all of the magnolia paint left by the previous owners) and next up is the landing outside our bedroom, which has become a bit of a dumping ground for clutter. Rachel made the excellent suggestion of a narrow table to sit on the landing and prevent clutter build-up, which would also form part of a feature wall with some art work. As well as a new piece to build after the Roubo bench is complete, this gave me a reason to start sketching.


Sketches 1 (top) and 2 (bottom) were exercises in seeing how the components might fit together.

I knew I wanted to try something different to my previous furniture builds, including greater use of curves, and a floating table top. That all sounds like a lot of fun, but how would it fit together? Would the legs go at the corners of the table top, or within the length? Would there be a lower stretcher, or just stretchers at the top of the legs? What would the proportions of the table be? I wasn’t thinking about joinery or technical solutions at this stage, just how the various elements that make a table might be arranged. I had no answers, so I started to draw, limiting myself to 60 seconds or so for each sketch. Now these are rough, rough drawings, but over the course of four sketches I started to zone in on how I thought the table should look – a curved tapered leg profile, with the legs situated underneath the table top allowing a good overhang at each end. A thin, straight stretcher towards the bottom of each leg, and a more substantial curved stretcher at the top. Possibly chamfering the ends of the table top to add to the curved feel.


Sketches 3 (top) and 4 (bottom) start to dial in how the design might look.

It didn’t take long to get to something I was broadly happy with. The next stage is to do a more detailed scale drawing (which is the design work I’m much more comfortable doing) and work out the specifics. And the notebook? The notebook will be a permanent resident in my work bag, and I’m setting myself a challenge to sketch furniture pieces at least 3 days a week for a month. At the end of the month, I’ll report back on whether that more regularised and focused design work has had any impact.

All About The Base?

Or: Roubo is Coming… part 8


Test fitting a long stretcher to the leg. That’s not a bad dry fit at all, and the small gap will close up when the joint is drawbored.

After chopping the remaining two mortises and tuning the fit of the tenons, the base of the workbench is now complete. While it would be tempting to dive straight into processing the slab there are a few other elements I want to attend to first, including fitting battens to the stretchers to support the shelf, and fitting the vise hardware, all of which will make life must easier when it comes to assembling the bench.

There is a lot of work left to do, but this feels like a huge milestone in the build process, and I am on track for having a completed bench before the year is out.