Robert updated the description to emphasize that the dirstate is corrupted by 'bzr mv' even before a subsequent commit.
But since the traceback in the comment on 2009-07-04 is actually generated in response to a second commit -- and because this is the context in which people are most likely to detect a problem (because bzr mv silently corrupts the dirstate) -- I have added some information to the end of the description.
Robert updated the description to emphasize that the dirstate is corrupted by 'bzr mv' even before a subsequent commit.
But since the traceback in the comment on 2009-07-04 is actually generated in response to a second commit -- and because this is the context in which people are most likely to detect a problem (because bzr mv silently corrupts the dirstate) -- I have added some information to the end of the description.