These days, we obsess over precise spoke length because we can. FB, Mr Rabbit, and other wheelbuilders with experience can probably relate to this posting by Ric Hjertberg.
https://www.wheelfanatyk.blogspot.ca/...e-lengths.html

I remember quite a few older folks on their way out back in the early to late 80s...

"10 and 7/8ths".

"Evertything four cross..."

"10mm nipples, old school long threads - cut off the rest when done. Who needs a calculator?"

And of course, retiring salesman would say: "You don't need a spoke machine..."

=8-)

When I started building 36h was the norm, except in the UK where they still preferred 32/40 F&R.

The nice thing about 36h, is that you could build them 4x. 4x pattern on a 36h is full tangent so flange diameter barely matters, and you could lace any wheel with typical rims on any hub with the same spokes. Dish was less then and you didn't even need to use two lengths on rears.

The brits also did something cool. By playing with flange size and cross they could build both the 32h front and 30h rear with the same spoke. That meant that a master wheel builder only needed to have a few size spokes on hand.

SS back then was rare and just starting to be trusted. Like Ric, we used mostly 15g DB Robergel spokes from France (different thread was a PIA if anybody should accidentally mix nipples). I scored a bunch of Stella 14g DB in stainless and am still using the last of them to this day.

As Ric said, we really had to get creative, and we learned to make do. My favorite make do solution was to use a 2mm drill to deepen the thread relief on nipples so they'd thread farther onto long spokes, a trick I still use when I have to.

Can't say I've ever resorted to that, and thankfully now with Mr. Morizumi in the house, I can trim off any amount and extend the threads, or apply 13mm of thread for long nipples.
Lucky are those like me who get to stand on the shoulders of the pioneers.

Very interesting. What does "full tangent" mean?

I means the spokes ran at right angles to imaginary lines from the axle center through the spoke holes in the flange and were. therefore. "tangent" to the edge of the flange.

Ahhhh! I bypassed the pain of compromising on spokes with all sorts of lacing patterns. I started building in the 80's and our shop had a Phil Wood spoke cutter. I remember reading the manual on how to line up the top part of the die to get the best and deepest rolled threads. And we had a bonanza for employees for a month when we cleared out our inventory of every damn spoke you could think of and started to stock just a few brands in max length to cut to order. The double-butted, we stocked in more lengths but still cut to order within a centimeter or so. Most of our employees couldn't do math or follow formulas worth jack. There was an insert in our Sutherland's binder that the owners added on spoke length calculation. I was one of just two of us that actually used it. The other guys would pull reference spokes out of the box and line them up against a busted wheel to guess what the spoke length would be and then use the cutter. Most of the time, it worked okay. We had an acceptable lost rate with a cup strapped to the side of the work bench where the Phil Wood cutter was with all the bad length cuts we made and we'd take those spokes first and cut ever shorter spokes. Thank goodness for 700C and 26 inch wheels because many of our repairs were for 27 inch wheels initially, and the transition allowed us to use those rejects and re-cut them. Only two of us actually could accurately build wheels from scratch and get the right spoke length. Even then, most of the wrench-heads thought the formula was magic.

Interestingly, the article pointed to by the OP mentions the HPV projects. I recall building the training wheels 4X 36 spokes sew-up rear and 20 inch aero clincher front that were used by our rider, the former Junior World Sprint Champion (who came to attend Cal Berkeley Engineering). Those were custom cut spokes I did in my shop. I wanted so much to be precise and then short by 1mm to save weight. But thought twice and didn't do it. I also thought about alloy nipples and butted spokes but decided that weight wasn't a concern. Better to train on heavy spoked straight spoked wheels and brass nipples, and our rider was a pretty big guy on a semi-recumbent that would be pretty brutal on the rear wheel. I followed the formula and ended up with just another wheelset like I had always built. Our actual racing wheel set was a custom aero wheels from Campagnolo only used in competition. We set the collegiate HPV land speed record back in the day at Sac-State on a 2-wheel semi-recumbent with faring. 3 years later, I went back to visit the Mech. Engineers who were still at it but this time sponsored. Campy, Dexter Hysol Adhesives, Boeing, Ford Aerospace... all giving $$$. They were still riding on those training wheels, and later set the HPV world speed record on their own with their jr. world sprint champion rider somewhere in the Sierra Foothills. They had always wanted to beat Fast Freddy Markham. But their record lasted for a short time until someone broke their land speed record. The Sac-State competition bike was called "Concept Z." The World Record bike was called Z-prime or something like that. I wonder where those training wheels went. They were "bomb" proof, they claimed and heavier than hell! LOL! I guess it made riding the racing wheels feel lighter when it came time. Wonder where those guys are now? And where are those training wheels? I think I donated them.

Back in the sixties, calculating spoke length by formula wasn't practical. There was no internet yet, and hand held scientific calculators were pricey toys that only engineering student could afford. The rest of us were left to struggle with pencil/paper, slide rules and sine tables.

Builders kept notes, so they never had to calculate spokes for the same wheel twice. I used a simple calculation method based on the distance of holes from the centerline, the rim OD, and the drop to the top of the nipple. Then added correction (fudge) factors for the flange separation or CTF for rears, and could generally hit my target within 1mm either way. I still use this method because it's simple and reliable.

As Ric noted in his blog, spoke size availability was thin, and back then there was no UPS, FEDEX, overnight shipping. You had to place large orders with local vendors who delivered once a week, or combine with bike orders for truck freight. Parts orders from non local vendors went Railway Express, or Greyhound and had to be picked up at the terminal.

The logistics alone, quickly trained bike mechanics in how to make do with what you had.

FBinNY, do you remember Larry Duffus? I think he was the first person I witnessed building a wheel. He would use his memory and educated guesses for spoke length. Then he'd lace the wheel up to discover that he was right or wrong. Then he'd go from 3x to 4x if the spokes were too long or vice versa if too short. It was a long, iterative process.

Yes, I knew Larry well both as a customer and friend. It's a shame that partner issues drove him away from the business. But his wheel bulding method was typical of old hands who predated the options available today. Until the Wheelsmith HP spoke calculator came on the market, spoke length calculations were a blend of good memories and the ability to guesstimate. Fortunately, back then rims didn't come in a million different depths, and consumers were more willing to listen to the builder than many are today.