The maximum practical number of nodes that can be connected to a 10BASE2 segment is limited to 30[4] with a minimum distance of 0.5 metres (20 in) between devices.
[b] If termination is missing, or if there is a break in the cable, the AC signal on the bus is reflected, rather than dissipated, when it reaches the end.
10BASE2 networks cannot generally be extended without breaking service temporarily for existing users and the presence of many joints in the cable also makes them very vulnerable to accidental or malicious disruption.
There were proprietary systems that claimed to avoid these problems (e.g. SaferTap) but these never became widespread, possibly due to a lack of standardization.
These characteristics made 10BASE2 ideal for a small network of two or three machines, perhaps in a home where easily concealed wiring was an advantage.
For a larger complex office network, the difficulties of tracing poor connections made it impractical.
Unfortunately for 10BASE2, by the time multiple home computer networks became common, the format had already been practically superseded by 10BASE-T. 10BASE2 uses RG-58A/U cable or similar for a maximum segment length of 185 m as opposed to the thicker RG-8-like cable used in 10BASE5 networks with a maximum length of 500 m. The RG-58 type wire used by 10BASE2 was inexpensive, smaller and much more flexible than the specialized RG-8 variant.
[6][7] An Ethernet network interface controller (NIC) may include the 10BASE2 transceivers and thus directly provide a 10BASE2 BNC connector (that the T-connector plugs into), or it may offer an AUI connector that external transceivers (see Medium Attachment Unit) can connect to.