?? 4d tangent cone of "2-rolls away" as containing 2d zariski tangent space of "1-roll-away" ... ??? does that allow us to take the 5d ambient tangent space and mod out by the 2d subspace, and have the 4d cone survive as a 2d cone ??? ..... ??? hmm, maybe not ... ?? could still try to project it down to the 3d quotient space, but .... hmm, it's only codimension 1 so pretty much any smearing is going to be total, right ?? .... ??? .... i'd been hoping that there was some way to get something here ... and that it would fit with how it seems very natural to get to _generic_ via roll-spinning ... ??? .... ?? had visual image of 1d line going through "middle" of cone .... ?? but now it seems like whole picture might be screwed up here .... ??? .....

?? maybe vague intuition here on right track ... ??? .... ?? "spin as proportionate to overspin" ... ??? if you know what i mean .... ??? ... ?? "yaw as proportionate to roll" .... ????? .... "proportionate" in some vague sense that should be made precise .... ??? ... ?? "yaw as proportionate to "difference" between roll and "speed"" ... ??? .... ?? so far still just trying to describe tangent cone .... ??? ....

?? actually i may have gotten "pitch" and "roll" here mixed up, for somewhat understandable reasons ... ?? but anyway, this raises the question of whether maybe both pitch and roll should enter into the situation somehow ..... ????? ...... ???? .....

?? "fixed points of tweak map" ??? .... ???? ....