For a system process error, agreed, we need show that only once per session: showing it more times wouldn't leave users any wiser.
For statistical purposes, we could aggregate subsequent errors from the same process in the same session, and report them together after the first system process error in any later session. A drawback is that it might result in errors being reported days after they happened, so that statistics for past days would need recalculating.
For a system process error, agreed, we need show that only once per session: showing it more times wouldn't leave users any wiser.
For statistical purposes, we could aggregate subsequent errors from the same process in the same session, and report them together after the first system process error in any later session. A drawback is that it might result in errors being reported days after they happened, so that statistics for past days would need recalculating.