2023-03-06 11 min read

Notes, 2023-03-06

Notes, 2023-03-06
The author, with a hacksaw, about to destroy a bicycle frame that he spent the previous month building. 2014.

One of the more interesting aspects of hanging out around young kids is watching them fail at something. Kids are constantly failing at something, and the ways in which they might misunderstand a problem (and misapply solutions to it) are surprising, and enlightening, and funny.

But as with a lot of parenting, watching my own kids fail can be terrifying. Not because of any imminent danger, but because managing, accepting, and responding to failure are all skills that I wish I was better at.

When someone fails at something over and over again, we call that practicing. It’s a word that my kids seem to have an innate repulsion to, and which I know I avoided as a child as well. To practice is to admit that you could improve, and also that you desire to improve – things I’m sure I was afraid to admit as a middle schooler, and which aren’t exactly natural to me now. But as I’ve gotten older, I’ve found that deliberate practice – failing at something over and over again with purpose – is one of the most gratifying things I can do.

-Spencer Wright

The most clicked link from last week's issue (5% of opens) was an essay on the importance and development of automobile paint. In the Members' Slack, we've been chatting about kettlebell logistics, using 3D printed lattices as RF lenses, and what's in store-bought chicken stock.


Ah, the number of things I’ve repeatedly failed at! Just this week, I’ve repeatedly failed to write this very newsletter – spending hours with the draft open, meandering from thought to introspective thought. Calling this “practice” feels like a rhetorical sleight of hand, but on days when I *don’t* do it, I’m left with a sense that I’m neglecting something.

In my experience, the life of a small business owner involves doing a *lot* of things that you are not good at. Most of my time is spent hovering right on the edge of failure on like four different work streams. To be fair, failure is relative to one’s own expectations, and I suppose the upside of small business ownership is that you get to decide what the business’s expectations are. Nevertheless, I don’t think I’m the first person to associate “running a small business” with “constantly being just a little bit out of your depth.” I suppose that the natural question is whether the repeated failures of a small business owner result in some form of mastery over time. But again, mastery is relative to one’s expectations, and I suspect that for most small business owners it is more interesting to have built something that doesn’t break down while you’re on vacation than it is to have developed any specific management or operational skill.

In his infamous 2009 essay titled Maker’s Schedule, Manager’s Schedule, Paul Graham suggested that “founders may increasingly be able to resist, or at least postpone, turning into managers, just as a few decades ago they started to be able to resist switching from jeans to suits.” The sentiment resonates with me, and recently I’ve begun to think that perhaps my pendulum has swung too far in the “chief dilettante” direction (a euphemism I use to describe my daily mix of activities) and not far enough in the “founding writer” direction. Then again, as Tom Vanderbilt explains in his heartwarming and energizing book Beginners, dilettante is derived from the Italian word dilettare to delight. And I suspect that I am not alone among small business owners who find some delight in their daily failures.

Three fillet brazed joints in various stages of finishing. 2010.


It’s been more than eleven years since I’ve done it on a regular basis, but the tactile feeling of cutting and finishing metal by hand is one that I can recall vividly. At the time I was building custom bicycle frames, and had decided for a few reasons (its relatively low risk; its niche nerd appeal) to use fillet brazing as my primary means of construction. Fillet brazing is a versatile method of joining metal, but it is used more in the high end bicycle industry than anywhere else that I’m aware of. To make a fillet brazed bicycle, you use a torch (typically oxygen-acetylene or oxygen-propane) to preheat your steel tubing and then melt and built up bronze filler rod on the joint. Once cooled, the finished joint is surprisingly strong: The Harris Group, which makes the popular low-fuming bronze filler rod that I mostly used, claims finished joint strengths around 480 MPa.

If you were just getting started building fillet brazed bicycle frames, you’d probably buy a bunch of 4130 steel tubing and start brazing chunks of it together to build up your skillset. Chances are, the tubes you buy will be marked “PLY/TUBE,” meaning that they were made by the Plymouth Tube Company, whose tubing is sold by a range of common industrial suppliers and marketed to motorsports and light aircraft manufacturers. They make tubing via cold drawing, meaning that the finished tube is made by pulling it really hard through a die (there’s a decent video of their process here; note that the tubes aren’t heated up during the drawing process). Specialty bicycle tubes (which are designed specifically for a particular location on the bicycle, a particular riding style, etc.) are made the same way, though they’re almost always butted or double-butted so that the wall thickness changes throughout the tube’s length. Of all the people I’ve spoken to in the bike industry over the years, Keith Noronha probably knows the most about specialty bicycle tubing; this interview with him, while not particularly riveting, has some nice footage of Reynolds tubing being butted.

When I was starting out, I’d cut (and cope) each tube using a hacksaw (a Bahco 325) and a series of mostly Swiss cut half-round files (like this one). Later I switched over to specialized coping fixtures mounted to a couple of old milling machines (primarily a fantastically burly Abene VHF3 universal, a Steinel SH4d horizontal, and a teeny little Benchmaster horizontal), but in retrospect I maintain a ton of appreciation for the hand work that I did early on. Hacksaws are cheap, versatile, and really quite good at cutting once you’ve learned how (smoothly) and when (on the push stroke) to apply pressure. Ditto with files, which inexperienced metalworkers will rub around aggressively but which are actually precise, effective, and satisfying to use once you’ve spent a couple dozen (hundred?) hours at the vise.Fillet brazing feels like TIG welding, but slowed down and zoomed in. You can see the process fairly well in this video – a single joint might take fifteen minutes to braze, after which it’s cooled and rinsed in warm water to dissolve the residual flux. Then, the hard finishing work begins. For seemingly arbitrary aesthetic reasons, the fillet on a fillet brazed bike is almost always ground, filed, and sanded until its surface is smooth and its edges lay tangent to the surface of the steel tubes. This process can take hours. Some framebuilders use a die grinder or handheld belt sander (both of which you can see here in the previous video) to finish their fillets, but I mostly used Swiss pattern files and rolled up pieces of sandpaper. It was grueling, and it destroyed my fingers, but the satisfaction of seeing two tubes blended together into a smooth joint was 😗👌

Early bicycle tires had a wide range of designs, few of which were compatible with each other. This design, described in 1906 in a trade book called Rubber Tires, was claimed to be popular because “the Darracq and Clement companies, now so famous as automobile makers, have generally controlled or rather monopolized French bicycle manufacture, and these manufacturers very early agreed to use only Dunlop tires on their wheels. This agreement had a powerful effect upon the independent French bicycle makers, who practically had to follow their lead and use Dunlops.”


My first real experience working on bikes was in college, in early 2003. The first bike I built up from parts was a steel road frame from the ‘90s that a previous owner had repainted a nice deep blue – leaving me clueless about who made it and unsure about the finer points of its geometry. In retrospect it was a size too big for me, but I invested in nice-ish (and at the time modern) components and rode it until I graduated in 2005, when I sold it on Craigslist along with an old recumbent I had picked up in the meantime. Since then, almost every new bike I’ve owned has been one that I built up from scratch – brazing or welding the frame myself, building the wheels by hand, and in many cases making some custom piece of hardware or another (a rack, a kickstand, a niche mechanical part). Some of this work was done when I was attempting to make a living as a custom bicycle framebuilder, but when I shut down my framebuilding business in 2011, I kept all of the bike tools. They now sit in a big tool chest behind my desk and are used on a more or less weekly basis.

In 2019, I wrote that “as one-off engineered objects – things that require design, fabrication, and systems integration – nothing beats a bicycle.” I stand by this claim, but I will also acknowledge that I derive a sense of moral superiority from my relationship with bikes. Bikes are so playful and earnest, and they simply overflow with my preferred signifiers of environmentalism and character-building. And anyway, my own sense of credibility relies on having at least one set of obscure (but useful) technical knowledge. Bikes provide me with that, and I have invested in my relationship with them accordingly. Today I own somewhere between six and ten, depending on how you count, and the likelihood that that number will decrease within my lifetime is relatively small.

Last week Kris De Decker, the author of Low-tech Magazine, published something of a takedown of newfangled bikes like mine, arguing that since the 1980s, “​​it becomes clear that the resource use of a bike's production increases while its lifetime is becoming shorter. The result is a growing environmental footprint.” De Decker’s 2018 post on the engineering behind his solar powered website is among my favorite articles on the internet, and when I observe it from afar, I can see some truth – or at least good intentions – in his perspective on the bicycle industry. Automobiles and airplanes are not the only carbon-emitting transportation forms out there, and it’s reasonable to proactively identify and reform the unsustainable aspects of the manufacturing processes we associate with environmental progress – things like bicycles, solar panels, and heat pumps.

But De Decker’s experience maintaining his bikes is quite different from mine, and I think a few of the claims he makes about modern bikes are simply false. So I tried to remember that his whole project – a valid one – is to underscore “the potential of past and often forgotten technologies.” And De Decker is right: Old bikes have potential. And whether his other conclusions are grounded in the bike industry that I know (and whether his suggestions are at all marketable to the general public) this core point is one that’s worth keeping in mind.

Low- to mid-range steel bike frames being welded at a small factory in Taiwan. By the time I toured this factory in 2014, I had spent years of my life trying to understand how to build steel bikes, building dozens of frames myself and visiting about a dozen small US framebuilders. The techniques, efficiency, and professionalism I saw in this small shop – which was totally unsophisticated, by the standards that the Taiwanese bike industry typically operates by – kind of blew my mind.


To a large extent, bicycles and bicycle components come from Taiwan – or are made by Taiwanese companies operating in China. Two manufacturers dominate bicycle manufacturing: Giant (whose founder, a failed eel farmer, was profiled by the NYTimes in 2013) and Merida (whose UK General Manager gave a relatively in-depth interview in 2022). Both of these companies were founded in 1972, and today they build a large portion of the bikes sold in the US – including most that are marketed under household names like Trek and Specialized. One Taiwanese company, Velo, absolutely dominates the market for bicycle saddles, making “roughly 80% of all enthusiast-level saddles sold worldwide.” I visited Velo’s factory in 2014 and met their founder, Stella Yu, who described her growth strategy in an interview with Cycling Tips two years later. She simply encouraged her customers to put their own brand names on the saddles that Velo made. “You [a bicycle brand] need to have your own exclusive product line. Now, all the most important brands have their own exclusive brands. That idea was from me. This is the Velo strategy.”

Stella Yu at the Velo factory showroom, 2014.

If you’re an independent bike shop owner, you probably buy partially disassembled bikes from one of a handful of name brands. Your mechanics do final assembly and tuning, and when they’re not building up new bikes (a task that typically goes to more junior mechanics) they’re doing service work. When a bike needs a replacement part, the most likely supplier will be QBP – Quality Bicycle Products, a Minnesota-based company that most established shops have an account with. (As a side note, QBP’s page on ImportYeti gives a good sense of where bike parts are made, and by who). Then there are BTI, HLC, JBI, and a handful of smaller distributors. Most of them will offer free shipping on orders over a few hundred dollars, making it easy for a shop to place an order weekly or even daily.


I love niche bicycle component manufacturers. The first one I ever visited in person was probably Phil Wood in San Jose, California (an old, low-fi tour of their shop is in this video). Later, in 2010, they made a few custom belt drive components for a bike I built for a customer.

Most of the bikes I own today have at least one part made by Paul Components (based and made in Chico, California; there’s a pretty good video tour here), and in an ideal world every bike I own would have titanium water bottle cages made by King Cage (who is based in Durango, Colorado, and whose fabrication process & shop setup are my *favorite*). I have a handful of components made by White Industries (based in Petaluma, California; they gave a video tour here) and a bunch of Thomson seatposts and stems (there’s a little manufacturing footage from their Macon, Georgia facility in this video), and a Spurcycle bell on basically anything I ride in traffic (who have a “built in the USA” video with some fabrication footage at the bottom of this page).

When I was in college, I saw these brands (all of which are owned and run by people who are more or less bike nerds like me) as separate from the bigger players in the industry. These days, though, I see the bike industry as being too small to support anyone who is separate – a feeling that’s probably related to the way in which I struggled to feel accepted by the bike industry myself. But that was so long ago now, and I was so bad at so many things then. And anyway, my sense of moral superiority – and my collection of six to ten bikes – will long outlive the bitterness I may have once felt about my repeated failures in the bike industry.


I’m generally on the lookout for techniques to manage my own stress, which can be triggered by a sense of futility – like when I fail at something a couple times in a row. Recently I’ve been toying with breathwork, doing daily five-minute “box breathing” exercises and trying “physiological sighs” when I’m particularly edgy. This study investigated those techniques’ real-time effects on stress, and the cyclic physiological sigh in particular seems to have done quite well.

Love, Spencer.

p.s. - We should be better friends. Send me a note - coffee's on me :)
p.p.s. - We care about inclusivity. Here's what we're doing about it.

Great! You’ve successfully signed up.
Welcome back! You've successfully signed in.
You've successfully subscribed to Scope of Work.
Your link has expired.
Success! Check your email for magic link to sign-in.
Success! Your billing info has been updated.
Your billing was not updated.