RFC 6920






Internet Engineering Task Force (IETF)                        S. Farrell
Request for Comments: 6920                        Trinity College Dublin
Category: Standards Track                                    D. Kutscher
ISSN: 2070-1721                                                      NEC
                                                            C. Dannewitz
                                                 University of Paderborn
                                                               B. Ohlman
                                                              A. Keranen
                                                                Ericsson
                                                         P. Hallam-Baker
                                                       Comodo Group Inc.
                                                              April 2013


                       Naming Things with Hashes

Abstract



   This document defines a set of ways to identify a thing (a digital
   object in this case) using the output from a hash function.  It
   specifies a new URI scheme for this purpose, a way to map these to
   HTTP URLs, and binary and human-speakable formats for these names.
   The various formats are designed to support, but not require, a
   strong link to the referenced object, such that the referenced object
   may be authenticated to the same degree as the reference to it.  The
   reason for this work is to standardise current uses of hash outputs
   in URLs and to support new information-centric applications and other
   uses of hash outputs in protocols.

Status of This Memo



   This is an Internet Standards Track document.

   This document is a product of the Internet Engineering Task Force
   (IETF).  It represents the consensus of the IETF community.  It has
   received public review and has been approved for publication by the
   Internet Engineering Steering Group (IESG).  Further information on
   Internet Standards is available in Section 2 of RFC 5741.

   Information about the current status of this document, any errata,
   and how to provide feedback on it may be obtained at
   http://www.rfc-editor.org/info/rfc6920.









Farrell, et al.              Standards Track                    [Page 1]

RFC 6920                Naming Things with Hashes             April 2013


Copyright Notice



   Copyright (c) 2013 IETF Trust and the persons identified as the
   document authors.  All rights reserved.

   This document is subject to BCP 78 and the IETF Trust's Legal
   Provisions Relating to IETF Documents
   (http://trustee.ietf.org/license-info) in effect on the date of
   publication of this document.  Please review these documents
   carefully, as they describe your rights and restrictions with respect
   to this document.  Code Components extracted from this document must
   include Simplified BSD License text as described in Section 4.e of
   the Trust Legal Provisions and are provided without warranty as
   described in the Simplified BSD License.

Table of Contents



   1.  Introduction . . . . . . . . . . . . . . . . . . . . . . . . .  3
   2.  Hashes Are What Count  . . . . . . . . . . . . . . . . . . . .  4
   3.  Named Information (ni) URI Format  . . . . . . . . . . . . . .  6
     3.1.  Content Type Query String Attribute  . . . . . . . . . . .  8
   4.  .well-known URI  . . . . . . . . . . . . . . . . . . . . . . .  9
   5.  URL Segment Format . . . . . . . . . . . . . . . . . . . . . . 10
   6.  Binary Format  . . . . . . . . . . . . . . . . . . . . . . . . 10
   7.  Human-Speakable (nih) URI Format . . . . . . . . . . . . . . . 11
   8.  Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
     8.1.  Hello World! . . . . . . . . . . . . . . . . . . . . . . . 13
     8.2.  Public Key Examples  . . . . . . . . . . . . . . . . . . . 13
     8.3.  nih Usage Example  . . . . . . . . . . . . . . . . . . . . 14
   9.  IANA Considerations  . . . . . . . . . . . . . . . . . . . . . 15
     9.1.  Assignment of ni URI Scheme  . . . . . . . . . . . . . . . 15
     9.2.  Assignment of nih URI Scheme . . . . . . . . . . . . . . . 15
     9.3.  Assignment of .well-known 'ni' URI . . . . . . . . . . . . 16
     9.4.  Creation of Named Information Hash Algorithm Registry  . . 16
     9.5.  Creation of Named Information Parameter Registry . . . . . 18
   10. Security Considerations  . . . . . . . . . . . . . . . . . . . 18
   11. Acknowledgments  . . . . . . . . . . . . . . . . . . . . . . . 20
   12. References . . . . . . . . . . . . . . . . . . . . . . . . . . 20
     12.1. Normative References . . . . . . . . . . . . . . . . . . . 20
     12.2. Informative References . . . . . . . . . . . . . . . . . . 21











Farrell, et al.              Standards Track                    [Page 2]

RFC 6920                Naming Things with Hashes             April 2013


1.  Introduction



   Identifiers -- names or locators -- are used in various protocols to
   identify resources.  In many scenarios, those identifiers contain
   values that are obtained from hash functions.  Different deployments
   have chosen different ways to include the hash function outputs in
   their identifiers, resulting in interoperability problems.

   This document defines the "Named Information" identifier, which
   provides a set of standard ways to use hash function outputs in
   names.  We begin with a few example uses for various ways to include
   hash function output in a name, with the specifics defined later in
   this document.  Figure 1 shows an example of the Named Information
   (ni) URI scheme that this document defines.

         ni:///sha-256;UyaQV-Ev4rdLoHyJJWCi11OHfrYv9E1aGQAlMO2X_-Q

                         Figure 1: Example ni URI

   Hash function outputs can be used to ensure uniqueness in terms of
   mapping URIs [RFC3986] to a specific resource or to make URIs hard to
   guess for security reasons.  Since there is no standard way to
   interpret those strings today, in general only the creator of the URI
   knows how to use the hash function output.  Other protocols, such as
   application-layer protocols for accessing "smart objects" in
   constrained environments, also require more compact (e.g., binary)
   forms of such identifiers.  In yet other situations, people may have
   to speak such values, e.g., in a voice call (see Section 8.3), in
   order to confirm the presence or absence of a resource.

   As another example, protocols for accessing in-network storage
   servers need a way to identify stored resources uniquely and in a
   location-independent way so that replicas on different servers can be
   accessed by the same name.  Also, such applications may require
   verification that a resource representation that has been obtained
   actually corresponds to the name that was used to request the
   resource, i.e., verifying the binding between the data and the name,
   which is here termed "name-data integrity".

   Similarly, in the context of information-centric networking
   [NETINF-ARCHITECTURE] [CCN] and elsewhere, there is value in being
   able to compare a presented resource against the URI that was used to
   access that resource.  If a cryptographically strong comparison
   function can be used, then this allows for many forms of in-network
   storage, without requiring as much trust in the infrastructure used
   to present the resource.  The outputs of hash functions can be used
   in this manner, if they are presented in a standard way.




Farrell, et al.              Standards Track                    [Page 3]

RFC 6920                Naming Things with Hashes             April 2013


   Additional applications might include creating references from web
   pages delivered over HTTP/TLS; DNS resource records signed using
   DNSSEC or data values embedded in certificates, Certificate
   Revocation Lists (CRLs), or other signed data objects.

   The Named Identifier can be represented in a number of ways: using
   the ni URI scheme that we specifically define for the name (which is
   very similar to the "magnet link" that is informally defined in other
   protocols [Magnet]), or using other mechanisms also defined herein.
   However it is represented, the Named Identifier *names* a resource,
   and the mechanism used to dereference the name and to *locate* the
   named resource needs to be known by the entity that dereferences it.

   Media content-type, alternative locations for retrieval, and other
   additional information about a resource named using this scheme can
   be provided using a query string.  "The Named Information (ni) URI
   Scheme: Optional Features" [DECPARAMS] describes specific values that
   can be used in such query strings for these various purposes and
   other extensions to this basic format specification.

   In addition, we define a ".well-known" URL equivalent, a way to
   include a hash as a segment of an HTTP URL, a binary format for use
   in protocols that require more compact names, and a human-speakable
   text form that could be used, e.g., for reading out (parts of) the
   name over a voice connection.

   Not all uses of these names require use of the full hash output --
   truncated hashes can be safely used in some environments.  For this
   reason, we define a new IANA registry for hash functions to be used
   with this specification so as not to mix strong and weak (truncated)
   hash algorithms in other protocol registries.

   The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
   "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this
   document are to be interpreted as described in RFC 2119 [RFC2119].

   Syntax definitions in this memo are specified according to ABNF
   [RFC5234].

2.  Hashes Are What Count



   This section contains basic considerations related to how we use hash
   function outputs that are common to all formats.

   When comparing two names of the form defined here, an implementation
   MUST only consider the digest algorithm and the digest value, i.e.,
   it MUST NOT consider other fields defined below (such as an authority
   field from a URI or any parameters).  Implementations MUST consider



Farrell, et al.              Standards Track                    [Page 4]

RFC 6920                Naming Things with Hashes             April 2013


   two hashes identical, regardless of encoding, if the decoded hashes
   are based on the same algorithm and have the same length and the same
   binary value.  In that case, the two names can be treated as
   referring to the same thing.

   The sha-256 algorithm as specified in [SHA-256] is mandatory to
   implement; that is, implementations MUST be able to generate/send and
   to accept/process names based on a sha-256 hash.  However,
   implementations MAY support additional hash algorithms and MAY use
   those for specific names, for example, in a constrained environment
   where sha-256 is non-optimal or where truncated names are needed to
   fit into corresponding protocols (when a higher collision probability
   can be tolerated).

   Truncated hashes MAY be supported.  When a hash value is truncated,
   the name MUST indicate this.  Therefore, we use different hash
   algorithm strings in these cases, such as sha-256-32 for a 32-bit
   truncation of a sha-256 output.  A 32-bit truncated hash is
   essentially useless for security in almost all cases but might be
   useful for naming.  With current best practices [RFC3766], very few,
   if any, applications making use of names with less than 100-bit
   hashes will have useful security properties.

   When a hash value is truncated to N bits, the leftmost N bits (that
   is, the most significant N bits in network byte order) from the
   binary representation of the hash value MUST be used as the truncated
   value.  An example of a 128-bit hash output truncated to 32 bits is
   shown in Figure 2.

                       128-bit hash: 0x265357902fe1b7e2a04b897c6025d7a2
              32-bit truncated hash: 0x26535790

                    Figure 2: Example of Truncated Hash

   When the input to the hash algorithm is a public key value, as may be
   used by various security protocols, the hash SHOULD be calculated
   over the public key in an X.509 SubjectPublicKeyInfo structure
   (Section 4.1 of [RFC5280]).  This input has been chosen primarily for
   compatibility with the DANE TSLA protocol [RFC6698] but also includes
   any relevant public key parameters in the hash input, which is
   sometimes necessary for security reasons.  This does not force use of
   X.509 or full compliance with [RFC5280] since formatting any public
   key as a SubjectPublicKeyInfo is relatively straightforward and well
   supported by libraries.

   Any of the formats defined below can be used to represent the
   resulting name for a public key.




Farrell, et al.              Standards Track                    [Page 5]

RFC 6920                Naming Things with Hashes             April 2013


   Other than in the aforementioned special case where public keys are
   used, we do not specify the hash function input here.  Other
   specifications are expected to define this.

3.  Named Information (ni) URI Format



   A Named Information (ni) URI consists of the following nine
   components:

   Scheme Name:  The scheme name is 'ni'.

   Colon and Slashes:  The literal "://"

   Authority:  The optional authority component may assist applications
      in accessing the object named by an ni URI.  There is no default
      value for the authority field.  (See Section 3.2.2 of [RFC3986]
      for details.)  While ni names with and without an authority differ
      syntactically from ni names with different authorities, all three
      refer to the same object if and only if the digest algorithm,
      length, and value are the same.

   One slash:  The literal "/"

   Digest Algorithm:  The name of the digest algorithm, as specified in
      the IANA registry defined in Section 9.4 below.

   Separator:  The literal ";"

   Digest Value:  The digest value MUST be encoded using the base64url
      [RFC4648] encoding, with no "=" padding characters.

   Query Parameter separator '?':  The query parameter separator acts as
      a separator between the digest value and the query parameters (if
      specified).  For compatibility with Internationalized Resource
      Identifiers (IRIs), non-ASCII characters in the query part MUST be
      encoded as UTF-8, and the resulting octets MUST be percent-encoded
      (see [RFC3986], Section 2.1).

   Query Parameters:  A "tag=value" list of optional query parameters as
      are used with HTTP URLs [RFC2616] with a separator character '&'
      between each.  For example, "foo=bar&baz=bat".

   It is OPTIONAL for implementations to check the integrity of the URI/
   resource mapping when sending, receiving, or processing ni URIs.

   Escaping of characters follows the rules in RFC 3986.  This means
   that percent-encoding is used to distinguish between reserved and
   unreserved functions of the same character in the same URI component.



Farrell, et al.              Standards Track                    [Page 6]

RFC 6920                Naming Things with Hashes             April 2013


   As an example, an ampersand ('&') is used in the query part to
   separate attribute-value pairs; therefore, an ampersand in a value
   has to be escaped as '%26'.  Note that the set of reserved characters
   differs for each component.  As an example, a slash ('/') does not
   have any reserved function in a query part and therefore does not
   have to be escaped.  However, it can still appear escaped as '%2f' or
   '%2F', and implementations have to be able to understand such escaped
   forms.  Also note that any characters outside those allowed in the
   respective URI component have to be escaped.

   The Named Information URI adapts the URI definition from the URI
   Generic Syntax [RFC3986].  We start with the base URI production:

         URI = scheme ":" hier-part [ "?" query ] [ "#" fragment ]
             ; from RFC 3986

                           Figure 3: URI Syntax

   Then, we adapt that for the Named Information URI:

         NI-URI         = ni-scheme ":" ni-hier-part [ "?" query ]
             ; adapted from "URI" in RFC 3986
             ; query is from RFC 3986, Section 3.4
         ni-scheme      = "ni"
         ni-hier-part   = "//" [ authority ] "/" alg-val
             ; authority is from RFC 3986, Section 3.2
         alg-val        = alg ";" val
             ; adapted from "hier-part" in RFC 3986
         alg            = 1*unreserved
         val            = 1*unreserved
             ; unreserved is from RFC 3986, Section 2.3

                         Figure 4: ni Name Syntax

   The "val" field MUST contain the output of base64url encoding (with
   no "=" padding characters), the result of applying the hash function
   ("alg") to its defined input, which defaults to the object bytes that
   are expected to be returned when the URI is dereferenced.

   Relative ni URIs can occur.  In such cases, the algorithm in Section
   5 of [RFC3986] applies.  As an example, in Figure 5, the absolute URI
   for "this third document" is "ni://example.com/sha-256-128;...".









Farrell, et al.              Standards Track                    [Page 7]

RFC 6920                Naming Things with Hashes             April 2013


     <html>
      <head>
        <title>ni: relative URI test</title>
        <base href="ni://example.com">
      </head>

      <body>
        <p>Please check <a href="sha-256;f4OxZX...">this document</a>.
          and <a href="sha-256;UyaQV...">this other document</a>.
          and <a href="sha-256-128;...">this third document</a>.
        </p>
      </body>
     </html>

                Figure 5: Example HTML with Relative ni URI

   The authority field in an ni URI is not quite the same as that from
   an HTTP URL, even though the same values (e.g., DNS names) may be
   usefully used in both.  For an ni URI, the authority does not control
   nearly as much of the structure of the "right-hand side" of the URI.
   With ni URIs we also define standard query string attributes and, of
   course, have a strictly defined way to include the hash value.

   Internationalisation of strings within ni names is handled exactly as
   for http URIs -- see [RFC2616], Section 3.2.3.

3.1.  Content Type Query String Attribute



   The semantics of a digest being used to establish a secure reference
   from an authenticated source to an external source may be a function
   of associated metadata such as the Content Type.  The Content Type
   "ct" parameter specifies the MIME Content Type of the associated data
   as defined in [RFC6838].  See Section 9.5 for the associated IANA
   registry for ni parameter names as shown in Figure 6.
   Implementations of this specification MUST support parsing the "ct="
   query string attribute name.

                   ni:///sha-256-32;f4OxZQ?ct=text/plain

                Figure 6: Example ni URI with Content Type

   Protocols making use of ni URIs will need to specify how to verify
   name-data integrity for the MIME Content Types that they need to
   process and will need to take into account possible Content-Transfer-
   Encodings and other aspects of MIME encoding.






Farrell, et al.              Standards Track                    [Page 8]

RFC 6920                Naming Things with Hashes             April 2013


   Implementations of this specification SHOULD support name-data
   integrity validation for at least the application/octet-stream
   Content Type, with no explicit Content-Transfer-Encoding (which is
   equivalent to binary).  Additional Content Types and Content-
   Transfer-Encodings can of course also be supported, but are OPTIONAL.
   Note that the hash is calculated after the Content-Transfer-Encoding
   is removed so it is applied to the raw data.

   If a) the user agent is sensitive to the Content Type and b) the ni
   name used has a "ct=" query string attribute and c) the object is
   retrieved (from a server) using a protocol that specifies a Content
   Type, then, if the two Content Types match, all is well.  If, in this
   situation, the Content Types do not match, then the client SHOULD
   handle that situation as a potential security error.  Content Type
   matching rules are defined in [RFC2045], Section 5.1.

4.  .well-known URI



   We define a mapping between URIs following the ni URI scheme and HTTP
   [RFC2616] or HTTPS [RFC2818] URLs that makes use of the .well-known
   URI [RFC5785] by defining an "ni" suffix (see Section 9).

   The HTTP(S) mapping MAY be used in any context where clients with
   support for ni URIs are not available.

   Since the .well-known name-space is not intended for general
   information retrieval, if an application dereferences a
   .well-known/ni URL via HTTP(S), then it will often receive a 3xx HTTP
   redirection response.  A server responding to a request for a
   .well-known/ni URL will often therefore return a 3xx response, and a
   client sending such a request MUST be able to handle that, as should
   any fully compliant HTTP [RFC2616] client.

   For an ni name of the form "ni://n-authority/alg;val?query-string"
   the corresponding HTTP(S) URL produced by this mapping is
   "http://h-authority/.well-known/ni/alg/val?query-string", where
   "h-authority" is derived as follows: If the ni name has a specified
   authority (i.e., the n-authority is non-empty), then the h-authority
   MUST have the same value.  If the ni name has no authority specified
   (i.e., the n-authority string is empty), a h-authority value MAY be
   derived from the application context.  For example, if the mapping is
   being done in the context of a web page, then the origin [RFC6454]
   for that web site can be used.  Of course, in general there are no
   guarantees that the object named by the ni URI will be available via
   the corresponding HTTP(S) URL.  But in the case that any data is
   returned, the retriever can determine whether or not it is content
   that matches the ni URI.




Farrell, et al.              Standards Track                    [Page 9]

RFC 6920                Naming Things with Hashes             April 2013


   If an application is presented with an HTTP(S) URL with
   "/.well-known/ni/" as the start of its pathname component, then the
   reverse mapping to an ni URI either including or excluding the
   authority might produce an ni URI that is meaningful.  However, there
   is no guarantee that this will be the case.

   When mapping from an ni URI to a .well-known URL, an implementation
   will have to decide between choosing an "http" or "https" URL.  If
   the object referenced does in fact match the hash in the URL, then
   there is arguably no need for additional data integrity, if the ni
   URI or .well-known URL was received "securely."  However, TLS also
   provides confidentiality, so there can still be reasons to use the
   "https" URL scheme even in this case.  Additionally, web server
   policy such as [RFC6797] may dictate that data only be available over
   "https".  In general, however, whether to use "http" or "https" is
   something that needs to be decided by the application.

5.  URL Segment Format



   Some applications may benefit from using hashes in existing HTTP URLs
   or other URLs.  To do this, one simply uses the "alg-val" production
   from the ni name scheme ABNF, which may be included, for example, in
   the pathname, query string, or even fragment components of HTTP URLs
   [RFC2616].  In such cases, there is nothing present in the URL that
   ensures that a client can depend on compliance with this
   specification, so clients MUST NOT assume that any URL with a
   pathname component that matches the "alg-val" production was in fact
   produced as a result of this specification.  That URL might or might
   not be related to this specification, only the context will tell.

6.  Binary Format



   If a more space-efficient version of the name is needed, the
   following binary format can be used.  The binary format name consists
   of two fields: a header and the hash value.  The header field defines
   how the identifier has been created, and the hash value contains a
   (possibly truncated) result of a one-way hash over whatever is being
   identified by the hash value.  The binary format of a name is shown
   in Figure 7.












Farrell, et al.              Standards Track                   [Page 10]

RFC 6920                Naming Things with Hashes             April 2013


      0                   1                   2                   3
      0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
     |Res| Suite ID  |              Hash Value                       /
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
     /                             ...                               /
     +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
     /      ...      |
     +-+-+-+-+-+-+-+-+

                       Figure 7: Binary Name Format

   The Res field is a reserved 2-bit field for future use and MUST be
   set to zero for this specification and ignored on receipt.

   The hash algorithm and truncation length are specified by the Suite
   ID.  For maintaining efficient encoding for the binary format, only a
   few hash algorithms and truncation lengths are supported.  See
   Section 9.4 for details.

   A hash value that is truncated to 120 bits will result in the overall
   name being a 128-bit value, which may be useful for protocols that
   can easily use 128-bit identifiers.

7.  Human-Speakable (nih) URI Format



   Sometimes a resource may need to be referred to via a name in a
   format that is easy for humans to read out and less likely to be
   ambiguous when heard.  This is intended to be usable, for example,
   over the phone in order to confirm the (current or future) presence
   or absence of a resource.  This "confirmation" use-case described
   further in Section 8.3 is the main current use-case for Named
   Information for Humans (nih) URIs. ("nih" also means "Not Invented
   Here", which is clearly false, and therefore worth including
   [RFC5513]. :-)

   The ni URI format is not well-suited for this, as, for example,
   base64url uses both uppercase and lowercase, which can easily cause
   confusion.  For this particular purpose ("speaking" the value of a
   hash output), the more verbose but less ambiguous (when spoken) nih
   URI scheme is defined.

   The justification for nih being a URI scheme is that it can help a
   user agent for the speaker to better display the value or help a
   machine to better speak or recognise the value when spoken.  We do
   not include the query string since there is no way to ensure that its
   value might be spoken unambiguously and similarly for the authority,
   where, e.g., some internationalised forms of domain name might not be



Farrell, et al.              Standards Track                   [Page 11]

RFC 6920                Naming Things with Hashes             April 2013


   easy to speak and comprehend easily.  This leaves the hash value as
   the only part of the ni URI that we feel can be usefully included.
   But since speakers or listeners (or speech recognition) may err, we
   also include a checkdigit to catch common errors and allow for the
   inclusion of "-" separators to make nih URIs easier to read out.

   Fields in nih URIs are separated by a semicolon (;) character.  The
   first field is a hash algorithm string, as in the ni URI format.  The
   hash value is represented using lowercase ASCII hex characters; for
   example, an octet with the decimal value 58 (0x3A) is encoded as
   '3a'.  This is the same as base16 encoding as defined in RFC 4648
   [RFC4648] except using lowercase letters.  Separators ("-"
   characters) MAY be interspersed in the hash value in any way to make
   those easier to read, typically grouping four or six characters with
   a separator between.

   The hash value MAY be followed by a semicolon ';' then a checkdigit.
   The checkdigit MUST be calculated using Luhn's mod N algorithm (with
   N=16) as defined in [ISOIEC7812] (see also [Luhn]).  The input to the
   calculation is the ASCII hex-encoded hash value (i.e., "sepval" in
   the ABNF production below) but with all "-" separator characters
   first stripped out.  This maps the ASCII hex so that '0'=0, ...'9'=9,
   'a'=10, ...'f'=15.  None of the other fields, nor any "-" separators,
   are input when calculating the checkdigit.

          humanname  = "nih:" alg-sepval [ ";" checkdigit ]
          alg-sepval = alg ";" sepval
          sepval     = 1*(ahlc / "-")
          ahlc       =  DIGIT / "a" / "b" / "c" / "d" / "e" / "f"
                ; DIGIT is defined in RFC 5234 and is 0-9
          checkdigit = ahlc

                     Figure 8: Human-Speakable Syntax

   For algorithms that have a Suite ID reserved (see Figure 11), the alg
   field MAY contain the ID value as an ASCII-encoded decimal number
   instead of the hash name string (for example, "3" instead of
   "sha-256-120").  Implementations MUST be able to match the decimal ID
   values for the algorithms and hash lengths that they support, even if
   they do not support the binary format.

   There is no such thing as a relative nih URI.









Farrell, et al.              Standards Track                   [Page 12]

RFC 6920                Naming Things with Hashes             April 2013


8.  Examples



8.1.  Hello World!



   The following ni URI is generated from the text "Hello World!" (12
   characters without the quotes), using the sha-256 algorithm shown
   with and without an authority field:

   ni:///sha-256;f4OxZX_x_FO5LcGBSKHWXfwtSx-j1ncoSt3SABJtkGk

   ni://example.com/sha-256;f4OxZX_x_FO5LcGBSKHWXfwtSx-j1ncoSt3SABJtkGk

   The following HTTP URL represents a mapping from the previous ni name
   based on the algorithm outlined above.

   http://example.com/.well-known/ni/sha-256/
   f4OxZX_x_FO5LcGBSKHWXfwtSx-j1ncoSt3SABJtkGk

8.2.  Public Key Examples



   Given the DER-encoded SubjectPublicKeyInfo in Figure 9, we derive the
   names shown in Figure 10 for this value.

          0000000 30 82 01 22 30 0d 06 09 2a 86 48 86 f7 0d 01 01
          0000020 01 05 00 03 82 01 0f 00 30 82 01 0a 02 82 01 01
          0000040 00 a2 5f 83 da 9b d9 f1 7a 3a 36 67 ba fd 5a 94
          0000060 0e cf 16 d5 5a 55 3a 5e d4 03 b1 65 8e 6d cf a3
          0000100 b7 db a4 e7 cc 0f 52 c6 7d 35 1d c4 68 c2 bd 7b
          0000120 9d db e4 0a d7 10 cd f9 53 20 ee 0d d7 56 6e 5b
          0000140 7a ae 2c 5f 83 0a 19 3c 72 58 96 d6 86 e8 0e e6
          0000160 94 eb 5c f2 90 3e f3 a8 8a 88 56 b6 cd 36 38 76
          0000200 22 97 b1 6b 3c 9c 07 f3 4f 97 08 a1 bc 29 38 9b
          0000220 81 06 2b 74 60 38 7a 93 2f 39 be 12 34 09 6e 0b
          0000240 57 10 b7 a3 7b f2 c6 ee d6 c1 e5 ec ae c5 9c 83
          0000260 14 f4 6b 58 e2 de f2 ff c9 77 07 e3 f3 4c 97 cf
          0000300 1a 28 9e 38 a1 b3 93 41 75 a1 a4 76 3f 4d 78 d7
          0000320 44 d6 1a e3 ce e2 5d c5 78 4c b5 31 22 2e c7 4b
          0000340 8c 6f 56 78 5c a1 c4 c0 1d ca e5 b9 44 d7 e9 90
          0000360 9c bc ee b0 a2 b1 dc da 6d a0 0f f6 ad 1e 2c 12
          0000400 a2 a7 66 60 3e 36 d4 91 41 c2 f2 e7 69 39 2c 9d
          0000420 d2 df b5 a3 44 95 48 7c 87 64 89 dd bf 05 01 ee
          0000440 dd 02 03 01 00 01

          0000000 53 26 90 57 e1 2f e2 b7 4b a0 7c 89 25 60 a2 d7
          0000020 53 87 7e b6 2f f4 4d 5a 19 00 25 30 ed 97 ff e4



             Figure 9: A SubjectPublicKeyInfo Used in Examples
                           and Its sha-256 Hash



Farrell, et al.              Standards Track                   [Page 13]

RFC 6920                Naming Things with Hashes             April 2013


   +-------------------------------------------------------------------+
   | URI:                                                              |
   | ni:///sha-256;UyaQV-Ev4rdLoHyJJWCi11OHfrYv9E1aGQAlMO2X_-Q         |
   +-------------------------------------------------------------------+
   | .well-known URL (split over 2 lines):                             |
   | http://example.com/.well-known/ni/sha256/                         |
   | UyaQV-Ev4rdLoHyJJWCi11OHfrYv9E1aGQAlMO2X_-Q                       |
   +-------------------------------------------------------------------+
   | URL Segment:                                                      |
   | sha-256;UyaQV-Ev4rdLoHyJJWCi11OHfrYv9E1aGQAlMO2X_-Q               |
   +-------------------------------------------------------------------+
   | Binary name (ASCII hex encoded) with 120-bit truncated hash value |
   | which is Suite ID 0x03:                                           |
   | 0353 2690 57e1 2fe2 b74b a07c 8925 60a2                           |
   +-------------------------------------------------------------------+
   | Human-speakable form of a name for this key (truncated to 120 bits|
   | in length) with checkdigit:                                       |
   | nih:sha-256-120;5326-9057-e12f-e2b7-4ba0-7c89-2560-a2;f           |
   +-------------------------------------------------------------------+
   | Human-speakable form of a name for this key (truncated to 32 bits |
   | in length) with checkdigit and no "-" separators:                 |
   | nih:sha-256-32;53269057;b                                         |
   +-------------------------------------------------------------------+
   | Human-speakable form using decimal presentation of the            |
   | algorithm ID (sha-256-120) with checkdigit:                       |
   | nih:3;532690-57e12f-e2b74b-a07c89-2560a2;f                        |
   +-------------------------------------------------------------------+

                         Figure 10: Example Names

8.3.  nih Usage Example



   Alice has set up a server node with an RSA key pair.  She uses an ni
   URI as the name for the public key that corresponds to the private
   key on that box.  Alice's node might identify itself using that ni
   URI in some protocol.

   Bob would like to believe that it's really Alice's node when his node
   interacts with the network and asks his friend Alice to tell him what
   public key she uses.  Alice hits the "tell someone the name of the
   public key" button on her admin user interface and that displays the
   nih URI and says "tell this to your buddy".  She phones Bob and reads
   the nih URI to him.








Farrell, et al.              Standards Track                   [Page 14]

RFC 6920                Naming Things with Hashes             April 2013


   Bob types that in to his "manage known nodes" admin application (or
   lets that application listen to part of the call), which can
   regenerate the ni URI and store that or some equivalent.  Then when
   Bob's node interacts with Alice's node, it can more safely accept a
   signature or encrypt data to Alice's node.

9.  IANA Considerations



9.1.  Assignment of ni URI Scheme



   The procedures for registration of a URI scheme are specified in RFC
   4395 [RFC4395].  The following assignment has been made.

   URI scheme name: ni

   Status: Permanent

   URI scheme syntax: See Section 3.

   URI scheme semantics: See Section 3.

   Encoding considerations: See Section 3.

   Applications/protocols that use this URI scheme name:

      General applicability.

   Interoperability considerations: Defined here.

   Security considerations: See Section 10.

   Contact: Stephen Farrell, stephen.farrell@cs.tcd.ie

   Author/Change controller: IETF

   References: As specified in this document

9.2.  Assignment of nih URI Scheme



   The procedures for registration of a URI scheme are specified in RFC
   4395 [RFC4395].  The following assignment has been made.

   URI scheme name: nih

   Status: Permanent

   URI scheme syntax: See Section 7.




Farrell, et al.              Standards Track                   [Page 15]

RFC 6920                Naming Things with Hashes             April 2013


   URI scheme semantics: See Section 7.

   Encoding considerations: See Section 7.

   Applications/protocols that use this URI scheme name:

      General applicability.

   Interoperability considerations: Defined here.

   Security considerations: See Section 10.

   Contact: Stephen Farrell, stephen.farrell@cs.tcd.ie

   Author/Change controller: IETF

   References: As specified in this document

9.3.  Assignment of .well-known 'ni' URI



   The procedures for registration of a Well-Known URI entry are
   specified in RFC 5785 [RFC5785].  The following assignment has been
   made.

   URI suffix: ni

   Change controller: IETF

   Specification document(s): This document

   Related information: None

9.4.  Creation of Named Information Hash Algorithm Registry



   IANA has created a new registry for hash algorithms as used in the
   name formats specified here; it is called the "Named Information Hash
   Algorithm Registry".  Future assignments are to be made through
   Expert Review [RFC5226].  This registry has five fields: the suite
   ID, the hash algorithm name string, the truncation length, the
   underlying algorithm reference, and a status field that indicates if
   the algorithm is current or deprecated and should no longer be used.
   The status field can have the value "current" or "deprecated".  Other
   values are reserved for possible future definition.

   If the status is "current", then that does not necessarily mean that
   the algorithm is "good" for any particular purpose, since the
   cryptographic strength requirements will be set by other applications
   or protocols.



Farrell, et al.              Standards Track                   [Page 16]

RFC 6920                Naming Things with Hashes             April 2013


   A request to mark an entry as "deprecated" can be done by sending a
   mail to the Designated Expert.  Before approving the request, the
   community MUST be consulted via a "call for comments" of at least two
   weeks by sending a mail to the IETF discussion list.

   Initial values are specified below.  The Designated Expert SHOULD
   generally approve additions that reference hash algorithms that are
   widely used in other IETF protocols.  In addition, the Designated
   Expert SHOULD NOT accept additions where the underlying hash function
   (with no truncation) is considered weak for collisions.  Part of the
   reasoning behind this last point is that inclusion of code for weak
   hash functions, e.g., the MD5 algorithm, can trigger costly false
   positives if code is audited for inclusion of obsolete ciphers.  See
   [RFC6149], [RFC6150], and [RFC6151] for examples of some hash
   functions that are considered obsolete in this sense.

   The suite ID field ("ID") can be empty or can have values between 0
   and 63, inclusive.  Because there are only 64 possible values, this
   field is OPTIONAL (leaving it empty if omitted).  Where the binary
   format is not expected to be used for a given hash algorithm, this
   field SHOULD be omitted.  If an entry is registered without a suite
   ID, the Designated Expert MAY allow for later allocation of a suite
   ID, if that appears warranted.  The Designated Expert MAY consult the
   community via a "call for comments" by sending a mail to the IETF
   discussion list before allocating a suite ID.

       ID  Hash Name String     Value Length     Reference   Status
       0   Reserved
       1   sha-256              256 bits         [SHA-256]   current
       2   sha-256-128          128 bits         [SHA-256]   current
       3   sha-256-120          120 bits         [SHA-256]   current
       4   sha-256-96           96 bits          [SHA-256]   current
       5   sha-256-64           64 bits          [SHA-256]   current
       6   sha-256-32           32 bits          [SHA-256]   current
       32  Reserved

                       Figure 11: Suite Identifiers

   The Suite ID value 32 is reserved for compatibility with IPv6
   addresses from the Special Purpose Address Registry [RFC4773], such
   as Overlay Routable Cryptographic Hash Identifiers (ORCHIDs)
   [RFC4843].

   The referenced hash algorithm matching the Suite ID, truncated to the
   length indicated, according to the description given in Section 2, is
   used for generating the hash.  The Designated Expert is responsible
   for ensuring that the document referenced for the hash algorithm
   meets the "specification required" rule.



Farrell, et al.              Standards Track                   [Page 17]

RFC 6920                Naming Things with Hashes             April 2013


9.5.  Creation of Named Information Parameter Registry



   IANA has created a new registry entitled "Named Information URI
   Parameter Definitions".

   The policy for future assignments to the registry is Expert Review,
   and as for the ni Hash Algorithm Registry above, the Designated
   Expert is responsible for ensuring that the document referenced for
   the parameter definition meets the "specification required" rule.

   The fields in this registry are the parameter name, a description,
   and a reference.  The parameter name MUST be such that it is suitable
   for use as a query string parameter name in an ni URI.  (See
   Section 3.)

   The initial contents of the registry are:

   Parameter    Meaning                                       Reference
   -----------  --------------------------------------------  ---------
   ct           Content Type                                  [RFC6920]

10.  Security Considerations



   No secret information is required to generate or verify a name of the
   form described here.  Therefore, a name like this can only provide
   evidence for the integrity of the referenced object, and the proof of
   integrity provided is only as good as the proof of integrity for the
   name from which we started.  In other words, the hash value can
   provide a name-data integrity binding between the name and the bytes
   returned when the name is dereferenced using some protocol.

   Disclosure of a name value does not necessarily entail disclosure of
   the referenced object but may enable an attacker to determine the
   contents of the referenced object by reference to a search engine or
   other data repository or, for a highly formatted object with little
   variation, by simply guessing the value and checking if the digest
   value matches.  So, the fact that these names contain hashes does not
   protect the confidentiality of the object that was input to the hash.

   The integrity of the referenced content would be compromised if a
   weak hash function were used.  SHA-256 is currently our preferred
   hash algorithm; this is why we've added only SHA-256-based suites to
   the initial IANA registry.

   If a truncated hash value is used, certain security properties will
   be affected.  In general, a hash algorithm is designed to produce
   sufficient bits to prevent a 'birthday attack' collision occurring.
   Ensuring that the difficulty of discovering two pieces of content



Farrell, et al.              Standards Track                   [Page 18]

RFC 6920                Naming Things with Hashes             April 2013


   that result in the same digest with a work factor O(2^x) by brute
   force requires a digest length of 2x.  Many security applications
   only require protection against a second pre-image attack, which only
   requires a digest length of x to achieve the same work factor.
   Basically, the shorter the hash value used, the less security benefit
   you can possibly get.

   An important thing to keep in mind is not to make the mistake of
   thinking two names are the same when they aren't.  For example, a
   name with a 32-bit truncated sha-256 hash is not the same as a name
   with the full 256 bits of hash output, even if the hash value for one
   is a prefix of that for the other.

   The reason for this is that if an application treats these as the
   same name, then that might open up a number of attacks.  For example,
   if I publish an object with the full hash, then I probably (in
   general) don't want some other application to treat a name with just
   the first 32 bits of that as referring to the same thing, since the
   32-bit name will have lots of colliding objects.  If ni or nih URIs
   become widely used, there will be many cases where names will occur
   more than once in application protocols, and it'll be unpredictable
   which instance of the name would be used for name-data integrity
   checking, thus leading to threats.  For this reason, we require that
   the algorithm, length, and value all match before we consider two
   names to be the same.

   The fact that an ni URI includes a domain name in the authority field
   by itself implies nothing about the relationship between the owner of
   the domain name and any content referenced by that URI.  While a
   name-data integrity service can be provided using ni URIs, that does
   not in any sense validate the authority part of the name.  For
   example, there is nothing to stop anyone from creating an ni URI
   containing a hash of someone else's content.  Application developers
   MUST NOT assume any relationship between the registrant of the domain
   name that is part of an ni URI and some matching content just because
   the ni URI matches that content.

   If name-data integrity is successfully validated, and the hash is
   strong and long enough, then the "web origin" [RFC6454] for the bytes
   of the named object is really going to be the place from which you
   get the ni name and not the place from which you get the bytes of the
   object.  This appears to offer a potential benefit if using ni names
   for scripts included from a HTML page accessed via server-
   authenticated https, for example.  If name-data integrity is not
   validated (and it is optional) or fails, then the web origin is, as
   usual, the place from which the object bytes were received.
   Applications making use of ni names SHOULD take this into account in
   their trust models.



Farrell, et al.              Standards Track                   [Page 19]

RFC 6920                Naming Things with Hashes             April 2013


   Some implementations might mishandle ni URIs that include non-base64
   characters, whitespace, or other non-conforming strings, and that
   could lead to erroneously considering names to be the same when they
   are not.  An ni URI that is malformed in such ways MUST NOT be
   treated as matching any other ni URI.  Implementers need to check the
   behaviour of libraries for such parsing problems.

11.  Acknowledgments



   This work has been supported by the EU FP7 project SAIL.  The authors
   would like to thank SAIL participants to our naming discussions,
   especially Jean-Francois Peltier, for their input.

   The authors would also like to thank Carsten Bormann, Martin Duerst,
   Tobias Heer, Bjoern Hoehrmann, Tero Kivinen, Barry Leiba, Larry
   Masinter, David McGrew, Alexey Melnikov, Bob Moskowitz, Jonathan
   Rees, Eric Rescorla, Zach Shelby, and Martin Thomas for their
   comments and input to the document.  Thanks, in particular, to James
   Manger for correcting the examples.

12.  References



12.1.  Normative References



   [ISOIEC7812]
              ISO, "Identification cards -- Identification of issuers --
              Part 1: Numbering system", ISO/IEC 7812-1:2006,
              October 2006, <http://www.iso.org/iso/iso_catalogue/
              catalogue_tc/catalogue_detail.htm?csnumber=39698>.

   [RFC2045]  Freed, N. and N. Borenstein, "Multipurpose Internet Mail
              Extensions (MIME) Part One: Format of Internet Message
              Bodies", RFC 2045, November 1996.

   [RFC2119]  Bradner, S., "Key words for use in RFCs to Indicate
              Requirement Levels", BCP 14, RFC 2119, March 1997.

   [RFC2616]  Fielding, R., Gettys, J., Mogul, J., Frystyk, H.,
              Masinter, L., Leach, P., and T. Berners-Lee, "Hypertext
              Transfer Protocol -- HTTP/1.1", RFC 2616, June 1999.

   [RFC2818]  Rescorla, E., "HTTP Over TLS", RFC 2818, May 2000.

   [RFC3986]  Berners-Lee, T., Fielding, R., and L. Masinter, "Uniform
              Resource Identifier (URI): Generic Syntax", STD 66,
              RFC 3986, January 2005.





Farrell, et al.              Standards Track                   [Page 20]

RFC 6920                Naming Things with Hashes             April 2013


   [RFC4395]  Hansen, T., Hardie, T., and L. Masinter, "Guidelines and
              Registration Procedures for New URI Schemes", BCP 35,
              RFC 4395, February 2006.

   [RFC4648]  Josefsson, S., "The Base16, Base32, and Base64 Data
              Encodings", RFC 4648, October 2006.

   [RFC5234]  Crocker, D. and P. Overell, "Augmented BNF for Syntax
              Specifications: ABNF", STD 68, RFC 5234, January 2008.

   [RFC5280]  Cooper, D., Santesson, S., Farrell, S., Boeyen, S.,
              Housley, R., and W. Polk, "Internet X.509 Public Key
              Infrastructure Certificate and Certificate Revocation List
              (CRL) Profile", RFC 5280, May 2008.

   [RFC5785]  Nottingham, M. and E. Hammer-Lahav, "Defining Well-Known
              Uniform Resource Identifiers (URIs)", RFC 5785,
              April 2010.

   [RFC6838]  Freed, N., Klensin, J., and T. Hansen, "Media Type
              Specifications and Registration Procedures", BCP 13,
              RFC 6838, January 2013.

   [SHA-256]  NIST, "Secure Hash Standard", FIPS 180-3, October 2008,
              <http://csrc.nist.gov/publications/fips/fips180-3/
              fips180-3_final.pdf>.

12.2.  Informative References



   [CCN]      Jacobson, V., Smetters, D., Thornton, J., Plass, M.,
              Briggs, N., and R. Braynard, "Networking Named Content",
              Proceedings of the 5th international conference on
              Emerging networking experiments and technologies (CoNEXT
              '09), December 2009.

   [DECPARAMS]
              Hallam-Baker, P., Stradling, R., Farrell, S., Kutscher,
              D., and B. Ohlman, "The Named Information (ni) URI Scheme:
              Optional Features", Work in Progress, June 2012.

   [Luhn]     Wikipedia, "Luhn mod N algorithm", September 2011,
              <http://en.wikipedia.org/w/
              index.php?title=Luhn_mod_N_algorithm&oldid=449928878>.

   [Magnet]   Wikipedia, "Magnet URI scheme", March 2013,
              <http://en.wikipedia.org/w/
              index.php?title=Magnet_URI_scheme&oldid=546892719>.




Farrell, et al.              Standards Track                   [Page 21]

RFC 6920                Naming Things with Hashes             April 2013


   [NETINF-ARCHITECTURE]
              Dannewitz, C., Kutscher, D., Ohlman, B., Farrell, S.,
              Ahlgren, B., and M. Karl, "Network of Information (NetInf)
              - An information-centric networking architecture",
              Computer Communications, Volume 36, Issue 7, pages
              721-735, ISSN 0140-3664, 1 April 2013,
              <http://www.sciencedirect.com/science/article/pii/
              S0140366413000364>.

   [RFC3766]  Orman, H. and P. Hoffman, "Determining Strengths For
              Public Keys Used For Exchanging Symmetric Keys", BCP 86,
              RFC 3766, April 2004.

   [RFC4773]  Huston, G., "Administration of the IANA Special Purpose
              IPv6 Address Block", RFC 4773, December 2006.

   [RFC4843]  Nikander, P., Laganier, J., and F. Dupont, "An IPv6 Prefix
              for Overlay Routable Cryptographic Hash Identifiers
              (ORCHID)", RFC 4843, April 2007.

   [RFC5226]  Narten, T. and H. Alvestrand, "Guidelines for Writing an
              IANA Considerations Section in RFCs", BCP 26, RFC 5226,
              May 2008.

   [RFC5513]  Farrel, A., "IANA Considerations for Three Letter
              Acronyms", RFC 5513, April 1 2009.

   [RFC6149]  Turner, S. and L. Chen, "MD2 to Historic Status",
              RFC 6149, March 2011.

   [RFC6150]  Turner, S. and L. Chen, "MD4 to Historic Status",
              RFC 6150, March 2011.

   [RFC6151]  Turner, S. and L. Chen, "Updated Security Considerations
              for the MD5 Message-Digest and the HMAC-MD5 Algorithms",
              RFC 6151, March 2011.

   [RFC6454]  Barth, A., "The Web Origin Concept", RFC 6454,
              December 2011.

   [RFC6698]  Hoffman, P. and J. Schlyter, "The DNS-Based Authentication
              of Named Entities (DANE) Transport Layer Security (TLS)
              Protocol: TLSA", RFC 6698, August 2012.

   [RFC6797]  Hodges, J., Jackson, C., and A. Barth, "HTTP Strict
              Transport Security (HSTS)", RFC 6797, November 2012.





Farrell, et al.              Standards Track                   [Page 22]

RFC 6920                Naming Things with Hashes             April 2013


Authors' Addresses



   Stephen Farrell
   Trinity College Dublin
   Dublin,   2
   Ireland
   Phone: +353-1-896-2354
   EMail: stephen.farrell@cs.tcd.ie


   Dirk Kutscher
   NEC
   Kurfuersten-Anlage 36
   Heidelberg
   Germany
   EMail: kutscher@neclab.eu


   Christian Dannewitz
   University of Paderborn
   Paderborn
   Germany
   EMail: cdannewitz@googlemail.com


   Borje Ohlman
   Ericsson
   Stockholm  S-16480
   Sweden
   EMail: Borje.Ohlman@ericsson.com


   Ari Keranen
   Ericsson
   Jorvas  02420
   Finland
   EMail: ari.keranen@ericsson.com


   Phillip Hallam-Baker
   Comodo Group Inc.
   EMail: philliph@comodo.com









Farrell, et al.              Standards Track                   [Page 23]