Howzit .

SQL doesn't keep these changes in a log as such . You have to design a strategy that will keep the changes for you and when it occurs .

You can use SQL profiler , hence the performance impact - so becarefull what events and couters you choose and also if you can design something that will keep all the entries in your syscacheobjects joined to the sysprocesses table in master you should catch the offender .

I have tried to catch a bugger dropping indexes during prod hours by creating a trigger on the syscacheobjects system table but with no -avail ..so I am also in the situasion ..I will keep you informed if I make any progress ...

To all other SQL fundies out there , give your comments , there is two of us now that would like you input in this regard .