Sentry is configured to track and report frontend errors reliably
There's a few questions to answer in this one. We want to track errors reliably, but we need to work out what reliably means in this case.
Here's some things we could discuss, though the list is not exhaustive:
-
Should we catch all errors and log everything, or should we only catch certain things? -
How do we prevent exessive noise from things like browser plugins? Will inbound data filters help? -
Will error sampling help? If so, how do we set it? -
Will we be able to re-enable source maps to help with debugging and error grouping?
Edited by Sam Beckham