If the Slink-e encounters an error of a critical nature, it reports the error, stops operating and enters the resume state. This is done so that host can resynchronize with the Slink-e. (If for example a receive buffer overflow occurred and a character was lost, you wouldn't want the Slink-e to act on the remaining bytes out of context). Once the host issues the resume command, normal operation resumes. If the Slink-e is controlled correctly, critical errors that require resuming should never occur.