[Fab-user] I bet it would be pretty to do transaction semantics but it sure would be nice

[Matthew Wilson]

There's a possibility that halfway through a long set of tasks,
something might go wrong.  In a database, I can use some kind of
rollback so that the system is in a coherent state.

I don't see how I could do something like this with fabric (I'm not
saying this is fabric's fault).

How are other people handling error situations?  It seems like a run
job immediately aborts.  Would it make sense to support a callback
that executes on failure?

Matt

-- 
Matthew Wilson
address@hidden
http://tplus1.com