rfc1255.txt

来自「RFC 的详细文档！」· 文本 代码 · 共 1,403 行 · 第 1/4 页

Network Working Group                 The North American Directory Forum
Request for Comments: 1255                                September 1991
Obsoletes:  RFC 1218


                        A Naming Scheme for c=US

Status of this Memo

   This memo provides information for the Internet community.  It does
   not specify an Internet standard.  Distribution of this memo is
   unlimited.

Summary

   This RFC is a near-verbatim copy of a document, known as NADF-175,
   which has been produced by the North American Directory Forum (NADF).
   The NADF is a collection of organizations which offer, or plan to
   offer, public Directory services in North America, based on the CCITT
   X.500 Recommendations.  As a part of its charter, the NADF must reach
   agreement as to how entries are named in the public portions of the
   North American Directory.  NADF-175 represents the NADF's agreement
   in this area.

Table of Contents

   1 Introduction ..........................................    2
   2 Approach ..............................................    2
   2.1 Names and User-Friendliness .........................    3
   2.2 Choice of RDN Names .................................    3
   2.3 Outline of the Scheme ...............................    4
   3 The Naming Process ....................................    4
   3.1 Right-To-Use ........................................    4
   3.2 Registration ........................................    6
   3.3 Publication .........................................    6
   4 Structuring Objects ...................................    7
   4.1 The National Level ..................................    7
   4.2 The Regional Level ..................................    7
   4.3 The Local Level .....................................    9
   4.4 ADDMD Operators .....................................   10
   4.5 Summary of Structuring Objects ......................   11
   5 Entity Objects ........................................   12
   5.1 Organizations .......................................   12
   5.1.1 Kinds of Organizations ............................   12
   5.1.2 Modeling Organizations ............................   13
   5.2 Persons .............................................   14
   6 Listing Entities ......................................   15
   6.1 Organizations .......................................   15



NADF                                                            [Page 1]

RFC 1255                A Naming Scheme for c=US          September 1991


   6.2 Persons .............................................   16
   7 Usage Examples ........................................   17
   7.1 Organizations with National-Standing ................   17
   7.2 Organizations with Regional-Standing ................   18
   7.3 Organizations with Local-Standing ...................   19
   7.4 Organizations with Foreign-Standing .................   20
   7.5 Persons .............................................   21
   8 Bibliography ..........................................   22
   Appendix A: Revision History of this Scheme .............   22
   Security Considerations .................................   25
   Author's Address ........................................   25

                         A Naming Scheme for c=US
                    The North American Directory Forum
                  Supercedes: NADF-166, 143, 123, 103, 71
                               July 12, 1991

1.  Introduction

   Computer networks form the infrastructure between the users they
   interconnect, and networks are built on an underlying naming and
   numbering infrastructure, usually in the form of names and addresses.
   For example, some authority must exist to assign network addresses to
   ensure that numbering collisions do not occur.  This is of paramount
   importance for an environment which consists of multiple service
   providers.

2.  Approach

   It should be observed that there are several different naming
   universes that could be used in the Directory Information Tree (DIT).
   For example, geographical naming, community naming, political naming,
   organizational naming, and so on.  The choice of naming universe
   largely determines the difficulty in mapping a user's query into a
   series of Directory operations to find useful information.  Although
   it is possible to simultaneously support multiple naming universes
   with the DIT, this is likely to be unnatural.  As such, this scheme
   focuses on a single naming universe.

   The naming universe in this scheme is based on civil authority.  That
   is, it uses the existing civil naming infrastructure and suggests a
   (nearly) straight-forward mapping on the DIT.  An important
   characteristic is that entries can be listed wherever searches for
   them are likely to occur.  This implies that a single object may be
   listed as several separate entries.






NADF                                                            [Page 2]

RFC 1255                A Naming Scheme for c=US          September 1991


2.1.  Names and User-Friendliness

   It must be emphasized that there are two distinct concepts which are
   often confused when discussing a naming scheme:

           (1)   user-friendly naming:
                 a property of a Directory which allows users to easily
                 identity objects of interest; and,


           (2)   Distinguished Name:
                 the administratively assigned name for an entry in the
                 OSI Directory.

   It must be emphasized that Distinguished Names are not necessarily
   user-friendly names, and further, that user-friendly naming in the
   Directory is a property of the Directory Service, not of
   Distinguished Names.

2.2.  Choice of RDN Names

   The key aspect to appreciate for choice of RDNs is that they should
   provide a large name space to avoid collisions: the naming strategy
   must provide enough "real estate" to accommodate a large demand for
   Distinguished Names.  This is the primary requirement for RDNs.  A
   secondary requirement is that RDNs should be meaningful (friendly to
   people) and should not impede searching.

   However, it is important to understand that this second requirement
   can be achieved by using additional (non- distinguished) attribute
   values.  For example, if the RDN of an entry is

      organizationName is Performance Systems International

   then it is perfectly acceptable (and indeed desirable) to have other
   values for the "organizationName" attribute, e.g.,

      organizationName is PSI

   The use of these abbreviated names greatly aids searching whilst
   avoiding unnecessary Distinguished Name conflicts.

   In order to appreciate the naming scheme which follows, it is
   important to understand that wherever possible it leverages existing
   naming infrastructure.  That is, it relies heavily on non-OSI naming
   authorities which already exist.  Note that inasmuch as it relies on
   existing naming authorities, there is little chance that any "final"
   national decision could obsolete this scheme.  (Any naming scheme may



NADF                                                            [Page 3]

RFC 1255                A Naming Scheme for c=US          September 1991


   be subject to the jurisdiction of certain national agencies.  For
   example, the US State Department is concerned with any impact on US
   telecommunications treaty obligations.) To do so would require a
   national decision that disregards existing national and regional
   infrastructure, and establishes some entirely new and different
   national naming infrastructure.

2.3.  Outline of the Scheme

   The naming scheme is divided into four parts:

           (1)   a discussion of the right-to-use, registration, and
                 publication concepts;

           (2)   a discussion of objects with national, regional, local,
                 and foreign standing;

           (3)   a discussion of objects which may be listed at
                 national, regional, and local levels; and,

           (4)   a discussion of how RDNs are formed for listing entries
                 at each different level.

3.  The Naming Process

   There are three stages to the naming process.

3.1.  Right-To-Use

   First, a naming authority must establish the right-to-use for any
   name to be used, within the jurisdiction of the given naming
   authority.  Names that are used in public are generally constrained
   by public laws.  Names that are only used in private are a private
   matter.  We are primarily concerned here with public names because
   these are the names that are most interesting to enter into public
   directories where we can search for them.

   There is a global governmental/civil/organizational infrastructure
   already in place to name and number things like people, cars, houses,
   buildings and streets; localities like populated places, cities,
   counties, states, and countries; organizations like businesses,
   schools, and governments; and other entities like computers,
   printers, ports, routers, processes, files, filesystems, networks,
   management domains, and so on.  There are also naming (and numbering)
   authorities for various standards and for networks (e.g., ISO/IEC,
   CCITT, IANA) which depend on acceptance by their constituent
   communities for their authority.




NADF                                                            [Page 4]

RFC 1255                A Naming Scheme for c=US          September 1991


   This collective infrastructure is comprised of a very large number of
   authorities that we will call naming authorities.  Naming authorities
   tend toward hierarchical organization.  Parents have authority
   (granted by government) to choose the names of new-born children, the
   courts have authority to change a person's name, car makers have
   authority to name the models of cars they build (within the limits of
   trademarking law), and they are obligated to assign unique serial
   numbers to each car.  Cities assign names to their streets and
   districts, states assign city, county, and township names, and so on.
   State governments also assign names to "registered" organizations
   that operate under state charters, which in turn name their own
   suborganizations.  Cities and Counties license businesses to use
   their chosen (unambiguous) names "in association with" the city and
   county names.  Companies name and number the computers and
   communications devices they make and sell.  There are many many name
   spaces, some of which are subordinate to others, and some of which
   are independent.

   Public names must be "registered" in some "public record" to record
   the fact of the assignment of the right-to-use to specific "owners."
   In general, this is to prevent collisions of the right-to-use
   assignments in public shared name spaces.  For example, unique names
   given to corporations are registered by the state of incorporation.
   A request to use a new name for any corporation must not conflict
   with the name of any other corporation registered in the same state.
   The same applies for businesses licensed within cities and counties.

   Establishment of the right-to-use for a name is not a Directory
   Service.  The right-to-use for a name is always derived from some
   other (non-directory) source of authority because of the legal
   aspects of intellectual property rights which are entirely outside
   the scope of directory service specifications.  People and
   organizations attach great value to the names they are allowed to
   associate with their lives and businesses, and intellectual property
   law protects their interests with respect to these values.

   This is not to say that directory service designers and providers
   have no interest in the processes and procedures for establishment of
   the right-to-use for the names that will be entered into any
   directory.  Indeed, without a supply of rightfully-usable names,
   there cannot be any directory.  But, given an adequate supply of
   registered names, the directory service is not otherwise concerned.

   We should note here that some naming authorities must deal with name
   spaces that are shared among large communities (such as computer
   networks) in which collisions will occur among applicants for desired
   name assignments, while other name spaces (such as for given names of
   children in a family) are not shared outside the family.  Sharing is



NADF                                                            [Page 5]

RFC 1255                A Naming Scheme for c=US          September 1991


   always a problem, which has led to trademarking laws, business
   license laws, and so on.  Naming within organizations should be
   easier, because it is "in the family," so to speak.  Hierarchical
   naming schemes facilitate distribution of naming authority.

3.2.  Registration

   Second, a name may be bound (as a value) to some object attribute.

   Given the right to use a name, a Naming Authority, such as a family
   which has an inherited surname and, more or less, has the right to
   use any names it pleases for its children's given names, must bind
   selected names to selected object attributes (e.g., firstname=Einar).
   Note that this same name might also be used as the first name or
   middle name of other children, as long as each sequence of given
   names of each family member is distinguished (i.e., none are
   duplicates) within the family.  Wise families do not bind the same
   sequence of given names to more than one child.  Some avoid any
   multiple use of a single name.  Some use generational qualifiers to
   prevent parent-child conflicts.

   The Internet Domain Name System (DNS) names top level domains which
   are then free (within some technical limits) to chose and bind names
   to entries which are subordinate to a given named domain, and so
   forth down the DNS name tree.  The ISO/CCITT naming system serves the
   same purposes in other separate name spaces.

3.3.  Publication

   Third, after binding, a name must be advertised or published in some
   community if it is to be referenced by others.  If it is not
   advertised or published, then no one can refer to it.

   This publication stage is what the Directory Service is all about.
   The Directory contains entries for "listed" names (or numbers) that
   are bound to the attributes of the entries in the directory DIT.
   Historically speaking, the directory business is a subclass of the
   publishing business, serving to dereference names into knowledge
   about what they stand for.

   It is important to keep in mind that a directory "listing entry" is
   not a "registration" unless a particular segment of the directory
   also just happens to be the authoritative master register of some
   naming authority.  Registration and listing are very different
   service functions, though it is conceivable that they might be
   combined in a single DIT.

   For example, in the United States of America, each state name is



NADF                                                            [Page 6]

RFC 1255                A Naming Scheme for c=US          September 1991


   registered by the Congress by inclusion of the name in the
   legislation that "admits each State into the Union." Note however
   that the name is also then published in many places (such as on maps
   and in directories), while the master "register" is kept with the
   other original records of laws enacted by the Congress and signed by
   the President.  Also, the name is then entered (listed) in many
   directories, in association with the name "The United States of
   America." And so on down the civil naming tree, with entities named
   in each state, etc.  It is certainly not the case that the American