> >Do you still have a copy of the originally damaged Database? I believe a >closer look to it will reveal more corruption than the assumed.
I have the original database. What other tests could I do to look for evidence? It appears (so far) that the database I reconstructed from the dump (minus an index request) contains all the original data, but that's hard to prove. >I do however think you are on the right track with thinking that the >duplicated transaction (or failure of rollback or program error or mishandling >of a duplicate insert fail that cause the transaction to be either committed >twice or not failed correctly or resubmitted without clear prior failure or >success) is the culprit here, and I am fairly certain this can happen in the >setup described above as per the http://www.sqlite.org/howtocorrupt.html It's likely in this case that the apple/microsoft cooperation on SMB is ultimately at fault because it is breaking some guarantee that sqlite is depending on. Their file sharing is known to be crap. It's still worthwhile to try to get to the bottom of it as a preliminary step to fixing it.