A shiny app deployed on an EC2 instance that has the capacity to run multiple sessions from a single R session.
It appears that the recommended way of triggering a callback when a shiny session encounters an error is the
shiny.error passes no information about the error, thus it appears impossible to identify the user session that precipitated the error.
We hope to write shiny input error logs, saved session information, and a traceback on a per user basis for debugging purposes, but the lack of specific information about the error and in the absence of the session object from that user's session seemingly make it impossible to do this.
Is this indeed the case, should this be a feature request?