They tried doing yestersol's drive in a new way: they picked a destination point and told the rover to go there in "blind waypoint" mode. In this mode, the rover keeps more careful track of its location than it does during a pure blind drive, but it doesn't do the full autonav thing either (monitoring the surrounding terrain for hazards and avoiding them). Only the very last part of the drive -- less than a meter or so out of the 20m total -- was done in the regular "blind" mode. Not only should the blind-waypoint approach be more accurate than pure blind driving (because it automatically corrects for slippage), it also turned out to be 17% faster. So when possible -- when the terrain is sufficiently benign, and our drive postconditions are sufficiently generous -- we'll probably do our "blind" drives that way from now on.
Not much else happens -- I don't stay very long. I get my pix fix, and at last I get around to teaching the MI PUL how to work out the terrain shadowing himself, so they won't have to ask me. And that's it.