Network Working Group N. Popp
Request for Comments: 2972 RealNames Corporation
Category: Informational M. Mealling
Network Solutions
L. Masinter
AT&T Labs
K. Sollins
MIT
October 2000
Context and Goals for Common Name Resolution
Status of this Memo
This memo provides information for the Internet community. It does
not specify an Internet standard of any kind. Distribution of this
memo is unlimited.
Copyright Notice
Copyright (C) The Internet Society (2000). All Rights Reserved.
Abstract
This document establishes the context and goals for a Common Name
Resolution Protocol. It defines the terminology used concerning a
"Common Name" and how one might be "resolved", and establishes the
distinction between "resolution" and more elaborate search
mechanisms. It establishes some eXPected contexts for use of Common
Name Resolution, and the criteria for evaluating a sUCcessful
protocol. It also analyzes the various motivations that would cause
services to provide Common Name resolution for both public, private
and commercial use.
This document is intended as input to the formation of a Common Name
Resolution Protocol working group. Please send any comments to
cnrp-ietf@lists.internic.net. To review the mail archives, see
<http://lists.internic.net/archives/cnrp-ietf.Html>
1. Introduction
People often refer to things in the real world by a common name or
phrase, e.g., a trade name, company name, or a book title. These
names are sometimes easier for people to remember and enter than
URLs; many people consider URLs hard to remember or type.
Furthermore, because of the limited syntax of URLs, companies and
individuals are finding that the ones that might be most reasonable
for their resources are already being used elsewhere and therefore
unavailable. Common names are not URIs (Uniform Resource
Identifiers) in that they lack the syntactic structure imposed by
URIs; furthermore, unlike URNs, there is no requirement of uniqueness
or persistence of the association between a common name and a
resource. These common names are expected to be used primarily by
humans (as opposed to machine agents).
Common name "resolution" is a process of mapping from common names to
Internet resources; a Common Name Resolution Protocol (CNRP) is a
network protocol used in such a process.
A useful analogy for understanding the purpose and scope of common
names, and CNRP, are everyday (human language) dictionaries. These
cover a given language (namespace) -- perhaps a spoken language, or
some specific subset (e.g., technical terms, etc). Some dictionaries
give definitions, others give translations (e.g., to other
languages). Different entities publish dictionaries that cover the
same language -- e.g., Larousse and Collins can both publish French-
language dictionaries. Thus, the dictionary publisher is the analog
to the resolution service provider -- the service can provide a
value-add and build up name recognition for itself, but does not
impede other entities from providing definitions for precisely the
same strings in the language.
Services are arising that offer a mapping from common names to
Internet resources (e.g., as identified by a URI). These services
often resolve common name categories such as company names, trade
names, or common keyWords. Thus, such a resolution service may
operate in one or a small number of categories or domains, or may
expect the client to limit the resolution scope to a limited number
of categories or domains. For example, the phrase "Internet
Engineering Task Force" is a common name in the "organization"
category, as is "Moby Dick" in the book category. A single common
name may be associated with different data records, and more than one
resolution service is expected to exist. Any common name may be used
in any resolution service.
Two classes of clients of such services are being built: browser
improvements and web Accessible front-end services. Browser
enhancements modify the "open" or "address" field of a browser so
that a common name can be entered instead of a URL. Internet search
sites integrate common name resolution services as a complement to
search. In both cases, these may be clients of back-end resolution
services. In the browser case, the browser must talk to a service
that will resolve the common name. The search sites are accessed via
a browser. In some cases, the search site may also be the back-end
resolution service, but in others, the search site is a front-end to
a collection of back-end services.
This effort is about the creation of a protocol for client
applications to communicate with common name resolution services, as
exemplified in both the browser enhancement and search site
paradigms. Although the protocol's primary function is resolution,
it is intended to address the issues of internationalization,
authentication and privacy as well. Name resolution services are not
generic search services and thus do not need to provide complex
Boolean query, relevance ranking or similar capabilities. The
protocol is expected to be a simple, minimal interoperable core.
Mechanisms for extension will be provided, so that additional
capabilities can be added later.
Several other issues, while of importance to the deployment of common
name resolution services, are outside of the resolution protocol
itself and are not in the initial scope of the proposed effort.
These include discovery and selection of resolution service
providers, administration of resolution services, name registration,
name ownership, and methods for creating, identifying or insuring
unique common names.
2. Key Goals for a Common Name Resolution Protocol
The key deliverable is a protocol for parameterized resolution.
"Resolution" is defined as the retrieval of data associated (a
priori) with descriptors that match the input request.
"Parameterized" means the ability to have a multi-component
descriptor both as part of the query and the response. These
descriptors are attribute-value pairs. They are not required to
provide unique identification, therefore 0 or more records may be
returned to meet a specific input query. The protocol will define:
- client requests/server responses to identify the specific
parameters accepted and/or required on input requests
- client request/server responses to identify properties to be
returned in the result set
- expression of parameterized input query
- expression of result sets
- standard expression of error conditions
To avoid creating a general search protocol with unbounded
complexity, and to keep the protocol simple enough so that different
implementations will have similar behavior, the resolution protocol
should be limited to sub-string matches against parameter values. To
support full internationalization, UTF-8 encoding of strings and
sub-strings is preferred.
In addition, the working group should define one sample service based
on this protocol -- the resolution of so-called "common names", or
resolution of non-unique, registered strings to resource
descriptions.
3. CNRP goals
The goal of CNRP is to create a lightweight search protocol with a
simple query interface, with a focus on making the common case of
substring search with a single result most efficient. In addition,
efficient support for keyed value search is important. Each key is a
named meta property of the resource (e.g. category, language,
geographical region.). Some of these properties could be
standardized (e.g. the common name property). The goal is to support
partial specification of query parameters and even partial and fuzzy
matches on names. CNRP is intended to be simpler than LDAP for
simple applications.
Besides simplicity, the CNRP protocol should be consistent with
efficient implementation of a simple and intuitive user interface.
The emphasis on the common name as the common denominator to find a
wide range of resources reduces the UI to its minimal expression (the
user types a few words in a text box and presses enter).
CNRP should provide interoperability with multiple common name
databases (section 4 presents many examples of such databases). The
query interface should be extensible and customizable to the specific
needs of a specific type of resolution service. However, the need
for interoperability across databases and resolution services
combined with the need to ensure the scalability of search (across
millions of names from multiple providers) have lead this group to
consider the explicit requirement of supporting categories in CNRP.
This requirement is discussed further in section 5.
4. Example of common name namespaces
Commercial companies have already developed and deployed common name
resolution services such as RealNames (http://www.realnames.com) and
NetWords (http://www.netword.com). These commercial implementations
are mainly focused on trade names, such as company names, brands and
trademarks. These services constitute a concrete example of common
name namespaces implementation and are useful to understand the scope
of the CNRP effort.
CNRP is also directly targeted at Directory service providers. CNRP
is relevant to these services to increase their reach through
integration into larger Web sites such as the search portals. For
example, IAtlas has developed a directory service for businesses that
it distributes through its Web site and Inktomi. IAtlas could
immediately leverage CNRP to distribute their service through their
external distribution partners.
Directory services must not be confused with search engines.
Directory services use highly structured information to identify a
resource. This information is external to the actual resource and is
called metadata. In contrast, search engines mainly rely on the
content of the resource (e.g. the text of a Web page).
CNRP plays well with directory services that present a critical piece
of information about the resource in the form of a textual
identifier, a title or a terse description (the common name).
Numerous examples come instantly to mind: company names, book titles,
people names, songs, ISBNs, and social security numbers. In all
cases, the common name is the natural property for users to lookup
the resource. The common name is always simple and intuitive: it has
no syntax, it is multilingual, memorable and can often be guessed.
The following list is intended to put in prospective the wide range
of applications for CNRP:
- Business directories (SEC, NASDAQ, E*Trade, .). The resource is
company information (address, products, SEC filings, stock quotes,
etc.). The common name is the company name.
- White pages (BigFoot, WhoWhere, Switchboard, ...): The resource a
person (current address, telephone numbers, email addresses,
employer, ...). The common name is a last name, a telephone number
or an email address.
- E-commerce directories: The resource is a product for sale (car,
house, furniture, actually almost any type of consumption item).
The common name is a brand name or a description.
- Publishing directories: The resource is one of many things: a book,
a poem, a CD, an mp3 download. The common name is an ISBN, a song
title, an artist's name. The common name is typically the title of
a publication.
- Entertainment directories: The resource is an event (a movie, a
concert, a TV show). The common name is the name or a description
for the event, the movie title, a rock band name, a show.
- Yellow pages services: Here again, the resource can be very
diverse: a house for sale, a restaurant, a car dealership or other
type of establishment or service that can be found in the
traditional yellow pages. The common name can be a street address,
the name of a business, or a description.
- News feeds: The resource is a press article. The common name is the
headline.
- Vertical directories: the DNS TLD categories, the ISO country
codes.
5. Private and public namespaces
A set of common names within a category (books, news, businesses,
etc.) is called a common name "namespace". The term "namespace" only
refers to the set of names. It does not encompass the bindings or
associations between a name and data about the name (such as a
resource, identified by a URI). Such bindings might be created and
maintained by a common name resolution services. Resolution services
may create binding that are relevant for the type of service that
they offer.
It is useful to distinguish between "private" and "public"
namespaces. A namespace is private if owned by an authority that
controls the right to assign the names. A namespace is private even
if the right to assign those names is held by a neutral party.
A namespace is public when not controlled by any single authority or
resolution provider. Assignment of the names is distributed.
However, it is reasonable to expect that people who assign names will
tend to pick names that have a minimum of collisions. For some of
these namespaces, there will even be mechanisms to discourage
duplicate assignment, but all of them are inherently ambiguous.
Public namespaces are not controlled. Examples of public namespaces
are:
- Titles of books, movies, songs, poems, short stories, plays, or
compilations
- Place names
- Street names
- People's names
Because these namespaces are unbounded and open to any types of name
assignment, they will have scalability problems. To support these
namespaces, CNRP must provide at least one standard mechanism to
filter a large list of related results. A filtering mechanism must
allow the user to narrow the search further down to a smaller result
set, because the common name alone may not be enough.
One possible search filter is related to the notion of categories.
Because categories create a structure to organize named resources,
large resolution services are likely to support some sort of
categorization system (whether flat or hierarchical). Although
categories constitute an efficient search filter, defining standard
vocabularies for common name categories is beyond the scope of the
protocol design. The protocol design for CNRP should not require a
standardized taxonomy for categories in order to be effective. For
example, CNRP resolution could use free-form keywords; the end-user
would use these keywords as part of the query. Each service would
then be responsible for mapping the keywords to zero, one or many
categories in their own classification. The keywords would remain
classification independent and different services could use different
categorization schemes without compromising interoperability. It
would then be up to the service to provide its own mapping. For
example, let us assume that one namespace is resolving names under
the category: "Hobby & Interests > collecting > antique > books".
Assume that a second namespace has decided to organize the names of
similar resources under the classification: "Arts > Humanities >
Literature > History of Books and Printing > antiques". Although the
two taxonomies are different, a CNRP query specifying
category_keywords = "antique books" would allow each service to
identify the appropriate category. This mechanism may ensure that
the two result lists are small and coherent enough to be merged into
one unique result set. It is important to note that this approach
would work whether the classification is hierarchical or not.
Although this suggestion has merit, it is fair to say that it remains
unproven. In particular, it is unclear that the category_keywords
property would guarantee full interoperability across resolution
services. In any case, free form keywords for specifying categories
is just one of several possible ways of limiting the scope of a
query. Although the specific mechanisms are not agreed upon a this
time, CNRP will provide at least one standard mechanism for limiting
scope.
6. Distributors/integrators of common name resolution services
We anticipate two main categories of distributors for common
namespaces. The first category is made of the Web portals such as
search engines (Yahoo, MSN, Lycos, Infoseek, AltaVista, ...). A
common name resolution service will typically address only one very
specialized ASPect of search (company names or book titles or people
names, ..). This type of focused lookup service is a useful
complement to generic search. Hence, portals are likely to integrate
several types of common name services. CNRP solves the difficult
problem of integrating multiple external independent services within
one Web site. Today, the lack of standardization in performance
requirements and query interface leads to loose integration (co-
branded pages hosted on virtual domains) or maintenance problems
(periodical data dumps). CNRP is aimed at solving some of these
issues. CNRP facilitates the deployment of embedded services by
creating a common interface to all common name services.
The second category of distributors is made of the Web browser
companies. Netscape's smart browsing
(http://home.netscape.com/communicator/v4.5/index.html#smart) and
Microsoft's IE5 auto-search features
(http://www.microsoft.com/windows/Ie/Features/AutoSearch/default.asp)
demonstrate that the two dominant Web browser companies understand
the value of navigation and search from the command line of the
browser. It is very clear how this command line could be used as the
main user interface to common name resolution services through CNRP.
In many ways, it is actually the most natural user interface to
resolve a common name. For this strategic component of the browser's
user interface to remain truly open to all common name resolution
services, it is key that there exists a standard resolution protocol
(and a service discovery mechanism). CNRP will give users access to
the largest selection of services and providers and the ability to
select a specific resolution service over another. To preserve the
user from proprietary implementations, the existence of CNRP is a
prerequisite.
7. Example of cost recovery models for maintenance of namespaces
The following discussion of possible business models for common name
namespaces is intended to prove that they are commercially viable,
hence that CNRP will be used in the market place. This section
presents 5 different cost recovery models.
a. Licensing the lookup service
In such model, the owner of the database owner licenses the data
and the resolution service to a portal. This is a proven model.
For example, Looksmart (a directory service) recently licensed all
their data to MSN. Another possibility is to sell access to the
service directly to the user. For some vertical type of common
names service (e.g. patent search), it is also conceivable that a
specific type of users (e.g., lawyers) would be willing to pay for
accessing a precise resolution service.
b. Sharing revenue generated by banner advertising
In this model, the database owner licenses his infrastructure
(data and resolution service) to a portal. Prepaid banner ads are
placed on the result pages. The revenue is shared between the
resolution service provider and the portal that hosts the pages.
c. Selling the names (charge the customer a fee for subscribing a
name)
This is a proven business model as well (NSI, GOTO, RealNames,
Netword, for of the name has a large user reach (search engines
sell keywords for instance).
d. Value added service
Another model is to build a common name as a free added value
service in order to make a core service more compelling to users.
For example, Amazon.com could create a common name namespace of
book titles and make it freely available to its users. Amazon.com
would not make any money from the resolution service per se.
However, it would indirectly since the service would help the
users find hence buy more books from Amazon.com.
e. "Some-strings-attached" free names
A namespace may give users a name for free in exchange for
something else (capturing the user's profile that can be sold to
merchants, capturing the user's email address in order to send
advertising emails, etc.).
8. Security and Intellectual Property Rights Considerations
This document describes the goals of a system for multi-valued
Internet identifiers. This document does not discuss resolution;
thus questions of secure or authenticated resolution mechanisms are
out of scope. It does not address means of validating the integrity
or authenticating the source or provenance of Common Names. Issues
regarding intellectual property rights associated with objects
identified by the various Common Names are also beyond the scope of
this document, as are questions about rights to the databases that
might be used to construct resolvers.
9. Authors' Addresses
Larry Masinter
AT&T Labs
75 Willow Road
Menlo Park, CA 94025
Phone: +1 650 463 7059
EMail: LMM@acm.org
http://larry.masinter.net
Michael Mealling
Network Solutions
505 Huntmar Park Drive
Herndon, VA 22070
Phone: (770) 935-5492
Fax: (703) 742-9552
EMail: michaelm@netsol.com
Nicolas Popp
RealNames Corporation
2 Circle Star Way
San Carlos, CA 94070-1350
Phone: 1-650-298-5549
EMail: nico@realnames.com
Karen Sollins
MIT Laboratory for Computer Science
545 Technology Sq.
Cambridge, MA 02139
Phone: +1 617 253 6006
EMail: sollins@lcs.mit.edu
10. Full Copyright Statement
Copyright (C) The Internet Society (2000). All Rights Reserved.
This document and translations of it may be copied and furnished to
others, and derivative works that comment on or otherwise explain it
or assist in its implementation may be prepared, copied, published
and distributed, in whole or in part, without restriction of any
kind, provided that the above copyright notice and this paragraph are
included on all such copies and derivative works. However, this
document itself may not be modified in any way, such as by removing
the copyright notice or references to the Internet Society or other
Internet organizations, except as needed for the purpose of
developing Internet standards in which case the procedures for
copyrights defined in the Internet Standards process must be
followed, or as required to translate it into languages other than
English.
The limited permissions granted above are perpetual and will not be
revoked by the Internet Society or its successors or assigns.
This document and the information contained herein is provided on an
"AS IS" basis and THE INTERNET SOCIETY AND THE INTERNET ENGINEERING
TASK FORCE DISCLAIMS ALL WARRANTIES, EXPRESS OR IMPLIED, INCLUDING
BUT NOT LIMITED TO ANY WARRANTY THAT THE USE OF THE INFORMATION
HEREIN WILL NOT INFRINGE ANY RIGHTS OR ANY IMPLIED WARRANTIES OF
MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.
Acknowledgement
Funding for the RFCEditor function is currently provided by the
Internet Society.