Currently, when something goes wrong, tbot dumps a stack-trace and a tiny exception message. Even worse, often it is hard to decipher what actually went wrong and how to deal with this.
Instead, tbot should display clean error messages which tell at least the following information:
What exactly went wrong (root cause), as good as tbot can know this.
Whether this is a bug in user code or an unexpected problem on some remote machine.
What are common solutions to this problem/where can a user find more information.
I'm not sure about stack-traces: Making then off-by-default and configurable might be annoying when a problem only occurs rarely and an unlucky user didn't have stack-traces enabled during the run. Always showing the full trace is over-verbose in most situations.
Currently, when something goes wrong, tbot dumps a stack-trace and a tiny exception message. Even worse, often it is hard to decipher what actually went wrong and how to deal with this.
Instead, tbot should display clean error messages which tell at least the following information:
I'm not sure about stack-traces: Making then off-by-default and configurable might be annoying when a problem only occurs rarely and an unlucky user didn't have stack-traces enabled during the run. Always showing the full trace is over-verbose in most situations.