errata logo graphic

Found 2 records.

Status: Verified (2)

RFC6588, "A URN Namespace for ucode", April 2012

Source of RFC: IETF - NON WORKING GROUP

Errata ID: 3188

Status: Verified
Type: Technical

Reported By: Alfred Hoenes
Date Reported: 2012-04-12
Verifier Name: Barry Leiba
Date Verified: 2012-04-13

Section 2, pg.3 says:

   Declaration of syntactic structure:

   The structure of the namespace for 'ucode' using the hexadecimal
   representation of the identifier is as follows using ABNF [RFC5234].

   UCODE-URN = "urn:ucode:" ucode-name
   ucode-name = "_" ucode-number
   ucode-number = 1*ucode-value
   ucode-value = 32HEXDIG
   HEXDIG         =  DIGIT / "A" / "B" / "C" / "D" / "E" / "F"
   DIGIT          =  %x30-39
                  ; 0-9

It should say:

   Declaration of syntactic structure:

   The structure of the namespace for 'ucode' using the hexadecimal
   representation of the identifier is as follows using ABNF [RFC5234].

   UCODE-URN    = "urn:ucode:" ucode-name
   ucode-name   = "_" ucode-number
   ucode-number = 1*ucode-value
|  ucode-value  = 32UCHEXDIG
|  UCHEXDIG     = %x30-39 / 0x41-46   ; digits 0..9, uppercase A..F
|

Notes:

Note: The above clause is part of the 'ucode' URN Namespace
Registration Template, so the above correction needs
to be applied to the template archived at IANA as well.

Rationale: The maintainers of the namespace intended to admit
only uppercase letters in the hexadecimal representation,
in order to accomodate usage of assigned <ucode-value>s in
case-sensitive XML context; this is specified in other parts
of the RFC, but should be specified also in the formal definition.
According to the ABNF Standard, RFC 5234, string literals in ABNF
are explicitly case-insensitive (cf. page 5 of RFC 5234).


Errata ID: 3189

Status: Verified
Type: Technical

Reported By: Alfred Hoenes
Date Reported: 2012-04-12
Verifier Name: Barry Leiba
Date Verified: 2012-04-13

Section 2, pg.4 says:

   Rules for lexical equivalence:

   The entire UCODE-URN is case-sensitive.

   NOTE: This is an additional restriction imposed on the ucode
   namespace by the requirements of some major applications of ucode in
   existence.  Only capital "A", "B", "C", ..., "F" are allowed as part
   of hexadecimal characters.

It should say:

   Rules for lexical equivalence:

|  The Namespace-Specific String (NSS) in 'ucode' URNs
|  (i.e. the <ucode-name> in the ABNF) is case-sensitive.
|  So this namespace imposes no additional lexical equivalences
|  beyond what is specified in RFC 2141 (i.e., according to
|  RFC 2141, the "urn:ucode:" part is case-insensitive, the NSS
|  is not).

Notes:

Note: The above clause is part of the 'ucode' URN Namespace
Registration Template, so the above correction needs
to be applied to the template archived at IANA as well.

Rationale: The RFC text violates Section 5 of RFC 2141, which
specifies that the case-insensitivity of the URI Scheme ("URN")
and the URN Namespace ID (NID) cannot be overridden by a URN
Namespace registration.
It was the intent of the maintainers of the 'ucode' namespace
to follow RFC 2141, but the language in the RFC has happened
to indicate otherwise.

The correction of the ABNF recorded in Errata Note #3188 makes
the original NOTE superflous, since the corrected ABNF now
precisely specifies what this NOTE intended to superimpose on
the original ABNF in the RFC.


Report New Errata