When we get a crash on bugsnag, the code writes the stack to our own json, then when the app starts up, if the ktcrash key matches the event, and there's only one error, we simply overwrite the stack with our stack. That's almost certainly fine, but I am worried that maybe that's not precise enough. This ticket is a reminder to dig into this later. We could try to do a "fingerprint match", where from main up until objc2kotlin should match function pointers exactly. However, I don't know if that buys us anything.
When we get a crash on bugsnag, the code writes the stack to our own json, then when the app starts up, if the
ktcrashkey matches the event, and there's only one error, we simply overwrite the stack with our stack. That's almost certainly fine, but I am worried that maybe that's not precise enough. This ticket is a reminder to dig into this later. We could try to do a "fingerprint match", where frommainup untilobjc2kotlinshould match function pointers exactly. However, I don't know if that buys us anything.