You need to remember that from a strictly practical perspective, unless you're breaking things while in their stock configuration, the quest for improving on them is of academic interest only.

I'd have to be really bored before I'd start to tear down a fully functional wheel only to incorporate an improvement that in all likelihood would be unnoticeable to me, particularly if I had as many bikes as SB had to tinker with.

When I got myself a pair of ceramic coated rims I treated myself to a wheel build with all the bells and whistles. DT Alpine in 3X on the DS and DT Rev in radial on the NDS(heads-out). Topped it off with lacing it in lateral cross. Linseed oil as lube/threadlock. Didn't have any trouble tensioning the Revs. The Alpines were a snug fit in the flange. It does even out tension balance a bit when compared to a 3X/3X build.
If I hadn't had to redish when I finally switched fom 7-speed to 9-speed I probably still wouldn't have touched the spokes, despite hard touring use.

Then again a regular 3x/3x built with the same loving care would probably have done just as well....

Revolutions only seem to become an issue with tensioning when you get over 100 kgf, at least that's what I found building my commuer wheelset (DT Rev 32 spoke, 3 cross disc wheel up front, DT Rev NDS, Comp DS, 32 spoke, 3 cross rear). My rear NDS spokes only reached ~90 kgf (offset rim) and a little chain lube dripped onto the nipples cured any twisting. The front wheel required me to hold some spokes with pliers to finally get the tension up without twisting the spokes. I only realized how much spoke twisting I had when I first tried to ride the front wheel. Putting the front brake on hard caused a lot of spoke pinging (a little unnerving).

Someone pointed out to me that radial lacing can void warranties on some hubs. Since I planned to build with a King hub, I checked and it does with them. So I think I may to 3xDS and 3xNDS, but use db2.0/1.8 on DS and db1.8/1.6 on NDS.

I keep a couple of those black paper binders/clips on my stand, and sometimes I'll clip them onto the spoke mid-length, as I turn the nipple, so I can get an idea of how much wind-up I'm getting. It's good for me to do as a reference every now and then.

My spokes don't wind up because as the tension starts up I hold every one of them with pliars. I don't see any other way with Revolutions.

Clip them near the nipple for an indicator of total wind-up. Then overtighten by 1/16-1/8th turn per adjustment then back off that same amount to remove the wind-up. This is an overblown issue by inexperienced wheelbuilders. Another thing to consider is you can undo the wind-up at the very end. It's a fixed amount, the spoke doesn't continue to wind-up more and more with each tightening of the nipple.

Those are the spokes with the lowest tension, most flex, and most often break. This is why I try to go really high on the driveside rear tension.

Revolutions will twist A LOT. I know this :) Competitions are much easier to work with.

I would be worried about damaging them; nicking them badly with the pliers...maybe that's paranoid.

There's no limit to the mount a spoke will wind up. The resistance to twisting is proportional to the fourth power of the diameter, so a DT revo will twist a lot more than even a 1.6 mm spoke. I suspect that at least one of the unreliable wheels I've had was caused by a builder who twisted a couple of spokes past the yield point.
Inexperienced builders should stick with heavier spokes until they are confident they can manage twisting.The resistance to twisting is proportional to the fourth power of the diameter, so a DT revo will twist a lot more than even a 1.6 mm spoke.

em

That has a smaller effect than when the spoke unloads as it rolls through the bottom of the wheel. I suspect the real reason that my gear changer hits the spokes is that the whole wheels bends when I rock the bike when I'm climbing.

em

Half radial spoking leaves the non-drive side with less tesion, not more. It's just a geometry problem, and the only way to get more equal tension is to use more crosses on the non-drive side, but the effect is too small to worry about.
Using lighter gauge spokes on the non-drive side equalizes stress, not tension. It makes a better wheel because the heavier spokes take more of the load.

That's exactly right.

em

If they've mastered the basics there's no reason to be afraid of twisting with thin spokes. It's just something to be managed. A simple method is to lace the wheel completely and pre-tighten to minimal tension where it has shape. At this point there shouldn't be enough thread friction for twisting, and/or you can stress relieve the rim to allow the spokes to untwist naturally. Now spin the wheel and hold a waterproof felt tip marker so it just touches the passing spokes above the nipple. That'll leave a reference line on each spoke that can be referenced for the natural untwisted orientation of the spoke.

Spokes can then be untwisted either in batches as the wheel moves through tightening stages, or near the final stages as each nipple is turned.

The key words there being "if they've mastered the basiss..."

That doesn't work if the spokes are twisted past yield. The reason I got back into wheelbuiding was that I got a really bad wheel from a local builder, and I have always thought that the problem was that several spokes were twisted past the yield point. That guy had built a lot of wheels, but he didn't understand what little margin of error there is when tensioning a 10 speed wheel. He hadn't "mastered the basics..."

em

No. A radial lace heads-out will get the same effect as a cross lace with the NDS flange half a thickness closer to the center of the axle. That will do a bit for evening out the tension differences. If you don't believe this, run it through spocalc yourself.

Wrong about the first part, right about the second.

If you want to go that route I believe the proper engineering term is strain, tension by unit of surface area.

As long as we're using engineering terms, we might as well use them properly. Strain is elongation/area, the word for tension/area is Stress, which is apropos of using smaller section spokes on the left rear. For the same tension (determined by the right side tension, and amount of dish) we have a greater relative loading on the spoke, ideally matching that on the right.

There's absolutely zero way to twist a spoke past yield-limit. You obviously don't know how to do the calculations.

I was right about everything. Your original post was about a heads-out radial lacing, which does cause a little lower tension. Using more crosses on one side and fewer on the other does equalize tension a little because the more crosses requires longer spokes that have a more slightly more acute angle than the short spokes with fewer crosses. Stress is tension/area. Strain is elongation/length, and is proportional to stress within the elastic range. Either stress or strain would have been a correct term.

em, p.e.