This document discusses different approaches for framing at the data link layer. It describes byte-oriented protocols like BISYNC, PPP that use sentinel characters or byte stuffing to delineate frames. The byte count approach used in DDCMP is also covered. For bit-oriented protocols, HDLC is described in detail, including its use of start/end bit sequences and bit stuffing to recognize frame boundaries despite corruption. Protocols like LCP, PAP, CHAP used along with PPP are also summarized briefly.