I think I've finally been able to put my finger on a bug that not only cost me several hours troubleshooting tonight before it twigged, but several other times when I thought it was just me going mad ... I'm using a lab environment with FIM version 4.0.3558.2 I had incorrectly implemented an IAF sync rule for a boolean metaverse person attribute IsActive derived from the AD account's userAccountControl attribute. Basically I had got the boolean results mixed up - returning true when it should have been false. After some time I was convinced that the IAF custom expression into my IsActive metaverse attribute should indeed have been Eq(BitAnd(userAccountControl,2),0). However, no matter what I did, nor how many variations I could think of to say the same thing, I couldn't toggle the results. No matter what I did, my expression resolved to false for an active account (512) and true for an inactive account (514)!!! Eventually it occurred to me that there MIGHT be a bug, and I should delete and recreate the attribute flow ... and VOILA ... problem solved. Up until this point it didn't matter what expression I put in I was getting the same result as before. So ... can anyone else vouch for this problem (i.e. has this a bug report number?). The work-around for now is easy ... delete and recreate your IAF :). Bob Bradley, www.unifysolutions.net (FIMBob?)

OK - checking the FIM build article now ... Mine is the November 2010 build ... checking all the subsequent build notes up to the latest 4.0.3576.2 ... It's possible that my scenario is described by "issue 4: A change to an incoming Synchronization Rule that includes a constant flow or an expression is not recalculated correctly during a full synchronization" of the 4.0.3573.2 build, but there's nothing more about it so I can only try applying the patch. There's certainly enough other stuff in this build for me to get the latest patch regardless :). I will roll forward to the latest build now (4.0.3576.2) and post back my results soon ...Bob Bradley, www.unifysolutions.net (FIMBob?)

Bob, that fix is intended for exactly the issue you describe. I bugged it back in October. It shipped just after Christmas and does indeed fix the issue.

Thanks Paul ... have just deployed the patch and you can assume it worked unless I post back here :)Bob Bradley, www.unifysolutions.net (FIMBob?)