kapers

Does anyone in the industry have any OFFICIAL opinion on what to do when the component specifies a standard Radial Contact bearing and has no provision for preload adjustment but the spec is still the industry standard C3 fit bearing? Sort of what happened in the day with Hershey hubs when some guy sat at a bench in thier service department and had a box of bearings on his left that were C2 and a box of bearings on his right that were C4 to deal with the tolerance discrepancies of the hub shells. A modern example (not to name names) would be a company that makes hub shells for several 'brands' that offer mechanically identical hubs in anodised, painted and plain/cnc'd, but does not ensure the bearing surface is cleaned of excess paint or allow for any changes in preload when the bearing is pressed in to the shell. If not, would anyone like an article about retaining compounds and thier various applications? Good ol' Roger Durham at Bullseye had a great theory on why his hubs (with noticable end-play, once tightened in frame/fork) were more efficient due to the very fact that not all of the balls were in contact when loaded. I wish I could find the notes (was pre-internet) but it was an interesting read and I'd like the industry opinion as well.