?? trying to visualize a2 real cartan torus action on flag variety ... orbit decomposition ...

?? 3d vector space ... with usual visualization of standard coordinate axises ...

?? cartan torus as "3d aspect ratio adjustment" group ...

?? subjecting 2d subspace to particular aspect ratio adjustment while _not_ subjecting 1d subspace to same adjustment ??? .....

??? "grassmanian blow-down" .... ???? hmmm, so what _does_ for example g2 "one-roll-away" give on _flag variety" as opposed to grassmanian ??? .... ???? .....

?? "pick direction, then magnitude" ... ????

?? span of schubert varieties here ..... ????? ......

?? non-/toric blow-ups/downs as spans of a sort ... ??? (?? better than as co-spans ?? ... ?? ...) .... ??relationship to bit we tried working out somewhat long time ago ... birational geometry .... ??? non-/toric .... ??? ....

?? confusion between ... ??choice of which destabilizer comes first, and which grassmanian coordinate gets omitted to give the partial flag blowdown ... ??? ..... ?? natural to omit the first one ?? .... ???? .... omitting the second one as giving the fiber that gets blowndown ??? ????? ....

??? blowing up light cone apex .... ???? ....

???? incidence span, pulled back .... ????? ......

??confusion ....

cartan .... cartan ..... ????? ......

??? span from blowndown variety to fiber of blowup ????? ..... ??? sounds screwed up ... ??? ....

?? many projections ... ??? composite span .... ??? ..... ???? ....