NIC view of standard host names
Rules for naming computers on the network
Watson, R.W., "NIC view of standard host names", RFC 237, DOI 10.17487/RFC0237, October 1971, <https://www.rfc-editor.org/info/rfc237>.
@misc{rfc237,
series = {Request for Comments},
number = 237,
howpublished = {RFC 237},
publisher = {RFC Editor},
doi = {10.17487/RFC0237},
url = {https://www.rfc-editor.org/info/rfc237},
author = {R.W. Watson},
title = {{NIC view of standard host names}},
pagetotal = 1,
year = 1971,
month = oct,
} RFC text
Network Working Group Richard W. Watson
Request for Comments: 237 October 5, 1971
NIC: 7662
The Nic's View of Standard Host Names
The NIC strongly favors standardization of host names. People will
get used to the names appearing on documents coming from or through
the NIC and so we would like the ones used by NIC and the ones used
in Telnets to agree. In the absence of agreed standard names, we
have chosen some. We are not particularly partial to the ones we are
using, but do feel any standard naming scheme take into account most
of the following considerations.
(1) Names consist of two parts, SITE and HOST, preferably
separated by a delimiter such as a dash.
(2) Site and Host names have some mnemonic value, if possible not
involving the name of the computer as computers will change, but
the organization running it will remain.
(3) We strongly recommend at least 8 characters be allowed,
including one delimiter. The NIC now uses 12 characters maximum.
(4) Assuming 8 character names, we recommend that the site name
be a maximum of 6 characters and that Host names at a given site
be unique in the first six characters. Thus, with the delimiter,
sites desiring to only have 4 characters in their tables can
accept all characters typed by the user and truncate to 6
characters and still achieve uniqueness.
(5) Once the NWG has set a standard syntax and names for present
Hosts, the NIC should be responsible for maintaining the standard
and assigning names to new hosts and other groups involved in
Network dialogue. Besides hosts there are NIC stations and
individuals in organizations not on the Net involved in the
Network dialogue. The NIC needs to assign names to these Network
affiliations and will use the rules set by the NWG.
[This RFC was put into machine readable form for entry]
[into the online RFC archives by H�l�ne Morin, Viag�nie 10/99]
AI ELI5
On this page
§1The Need for Standard Names
RFC 237 explains why the Network Information Center (NIC) favors standardized host names on the ARPANET. Different sites use different names for the same computers, which confuses users who read documents from the NIC and then try to connect using names they learned elsewhere.
When each computer had lots of different names, people got confused about which computer was which.
§2The NIC's Naming Scheme
The NIC proposes a two-part name structure: a site name and a host name, separated by a delimiter such as a dash (e.g., MIT-MULTICS). Names should be meaningful, referring to the organization and its system rather than the computer model, since computers change but organizations persist.
The scheme recommends at least eight characters including the delimiter. The site name should be a maximum of six characters, and host names should be unique within their site in the first six characters. This allows sites with only four-character tables to accept user input and truncate it while maintaining uniqueness.
Names like "Stanford-AI" or "MIT-Multics" tell you both where the computer is and what job it does. The NIC likes at least eight letters so there is room for both parts.
§3Role of the NIC
The NIC proposes that once the Network Working Group sets a standard syntax and names, the NIC should be responsible for maintaining the standard and assigning names to new hosts. This includes assigning names to NIC stations and individuals in organizations not yet on the network who need to participate in network dialogue.
The NIC would keep a master list of all the names and make sure no two computers got the same name.