This content has been marked as final. Show 2 replies
One idea is that I doubt that is viable.
Attempting a one size fits all solution to exception handling is unlikely to work.
Now if there is a specific small problem domain to be addressed that might be doable.
Some general thoughts: Consider whether it makes sense to set up error categories (i.e., base classes) for (1) warnings (2) transient errors (ones that might go away on a retry) or (3) fatal errors (ones you're pretty sure will require human intervention). My personal philosophy is that it makes sense to declare different exception types only if I handle them differently, or if the error types themselves give information about the cause or location of the problem.