FW: [ebxml-dev]

From
Pim van der Eijk
Date
2009-08-07T07:55:00+00:00
ID
Thread
FW: [ebxml-dev]
FYI.



From: Pim van der Eijk 
[mailto:]
Sent: 07 August 2009 
09:20
To: ''; 'Moberg Dale'; 'David RR 
Webber (XML)'
Cc: ''; 
''
Subject: RE: [ebxml-dev] <PartyId 
type="?SWIFT BIC?">BBBBCCLL123<\PartyId>




Hello,
 
IANA assigns the URN top-level namespaces, and OASIS applied for the "urn:oasis" one and (at 
one point) defined "names:tc" branch so TCs could define their own special-purpose 
namespaces. The OASIS TC has 
defined a sub-branch to refer to ISO 6523 registered organizations.  Those 
organizations would need to apply to the registration holder for ISO 6523 
themselves separately (I believe 6523 is handled by the BSI).  It would 
probably be a good idea for the OASIS TC to also define a urn:oasis:names:tc:ebxml-cppa:partyid-type:iso15022 
branch for organizations classified according to ISO 
15022.  
 
The CEN workshop people that proposed the "iso6523" 
branch will need to register it with IANA themselves, and their CWA indicates 
this is their intention.  If you want to use something like urn:iso9362:1994 or urn:iso15022 then in theory someone should apply for such top 
level URN domains. 
 
However there is 
no mechanism to resolve URNs like there is for HTTP URLs, so in reality these 
URNs are just identifier strings.
 
There is indeed a 
lot of overlap in these registrations.  Companies are classified on 
multiple schemes, and those classification schemes are in turn classified based 
on multiple meta-registration schemes like 6523 and 15022.  And now we are 
discussing multiple ways of designating such meta-registration schemes.  
There is more than one way that leads to Rome 
..
 
Pim



From: Charles Kilkenny 
[mailto:]
Sent: 06 August 2009 
18:50
To: 'Moberg Dale'; 'David RR Webber (XML)'; 'Pim van der 
Eijk'
Cc: ; 

Subject: RE: [ebxml-dev] <PartyId 
type="?SWIFT BIC?">BBBBCCLL123<\PartyId>




Hi,

Also, looking at it 
from a banking and finance perspective … what to do about other identifiers 
issued by exchanges (http://www.iso15022.org/MIC/homepageMIC.htm) 
or by central securities depositories (http://www.iso15022.org/Data_Source_Schemes.pdf)? 
Currently, these party identifiers are represented by a 4 character scheme e.g. 
“CRST” = “Euroclear CREST” followed by the participant code. That is in ISO 
15022 speak. And, some issuers, e.g. messaging network operators, are not even 
represented as a DSS (ISO 15022 data source scheme). Also, I guess ISO 15022 DSS 
most probably overlaps other schemes.

Charles




From: Charles 
Kilkenny [mailto:]
Sent: 06 August 2009 18:24
To: 'Moberg Dale'; 'David RR Webber (XML)'; 
'Pim van der Eijk'
Cc: 
''; ''
Subject: RE: [ebxml-dev] <PartyId 
type="?SWIFT 
BIC?">BBBBCCLL123<\PartyId>

Hi Dale/Pim & 
David,

Many thanks for your 
responses. They have been really helpful.

I can’t help saying 
though that the following could give the computer repetitive 
strains:

1. 
urn:oasis:names:tc:ebxml-cppa:partyid-type:S.W.I.F.T:0021

2. 
urn:oasis:names:tc:ebxml-cppa:partyid-type:SocietyForWorldwideInterbankFinancialTelecommunicationS.W.I.F.T.:0021

This is a lot 
better:

3. 
urn:oasis:names:tc:ebxml-cppa:partyid-type:iso6523:0021

And, this would be even 
better:

4. 
urn:iso6523:0060

Option (4) would also 
remove “ebxml” (sorry all you ebXML folks) which would enable it to be used 
outside of ebXML and hence have even more global adoption J

There may just be one 
snag however and I’m not completely sure. I’m not sure whether SWIFT is just the 
registrar (I suspect though that this is the case) or whether it actually owns 
the BIC coding scheme. If the former, the following may be more 
appropriate:

5. 
urn:iso9362:1994

An authoritative Wiki 
as a central repository would also be great. I could then stop 
searching!

Thanks 
again.
Charles




From: Moberg 
Dale [mailto:]
Sent: 06 August 2009 17:51
To: David RR Webber (XML); Pim van der 
Eijk
Cc: Pim van der Eijk; 
; 
; 
Subject: RE: [ebxml-dev] <PartyId 
type="?SWIFT 
BIC?">BBBBCCLL123<\PartyId>

Let’s discuss tomorrow 
morning on ebCore call. Dale Moberg



(I think the iso6523 
allows short identifiers for all iso registered values. Not certain that iso is 
maintaining that code list however.)

Good idea to have a 
wiki. We could post all the known v2 and v 3 codes deriving from code 
registries.







From: David RR 
Webber (XML) [mailto:]
Sent: Thursday, August 06, 2009 7:46 
AM
To: Pim van der 
Eijk
Cc: 'Pim van der Eijk'; 
; ; 

Subject: RE: [ebxml-dev] <PartyId 
type="?SWIFT 
BIC?">BBBBCCLL123<\PartyId>


My suggest is to use 
the OASIS wiki to support this in future - simple to maintain and 
find.



We will have to 
request tc-admin to set ebcore up - http://wiki.oasis-open.org/ 



Then we could "fix" 
the broken link by a redirect to the appropriate wiki 
page.


Thanks, 
DW

  
  -------- Original 
  Message --------
Subject: RE: [ebxml-dev] <PartyId 
  type="?SWIFT
BIC?">BBBBCCLL123<\PartyId>
From: "Pim van der 
  Eijk" <>
Date: Wed, August 05, 2009 3:11 
  pm
To: "''" 
  <>,
<>, 
  <>
Cc: "'Pim van der Eijk'" 
  <>
  Dale and others,
   
  There was 
  some discussion on this recently on UBL-DEV:
  http://lists.oasis-open.org/archives/ubl-dev/200907/msg00012.html
   
  For SWIFT, 
  something like:
   
  urn:oasis:names:tc:ebxml-cppa:partyid-type:iso6523:0021
   
  would 
  work.  
   
  Pim
  
  
  
  
  From:  [mailto:]
Sent: 05 August 2009 
  17:40
To: ; 
  
Cc: Pim van der 
  Eijk
Subject: RE: [ebxml-dev] 
  <PartyId type="?SWIFT 
  BIC?">BBBBCCLL123<\PartyId>
  
  
  Hi,
  
  
  
  
  
  
  Thanks 
  for pointing out the broken link. I will ask OASIS whether they can fix it. 
  The ebCore TC now is charged with maintenance of CPPA 
  versions.
  
  
  
  The 
  version 3 of CPPA is progressing slowly and will have the approved party id 
  identifier material included whenever we are able to finish the editorial 
  work.
  
  
  
  Meanwhile the material from version 3, 
  section 4.3.1 (subject to change) that should also be at the link is attached 
  to this reply. It is in the open office open document format. (There may be 
  some variation from the material at the link but I cannot be certain until I 
  get OASIS to fix the link.)
  
  
  
  [Also 
  the text is short enough to insert after discussion of the SWIFT 
  value.]
  
  
  
   Now for Swift from 
  ISO 6523 Data 
  interchange- Structure for the identification of organizations, I 
  find
  
  
  
  ICD : 
  0021
  
  Name 
  of coding  system: SOCIETY FOR WORLDWIDE INTERBANK 
  FINANCIAL
  
   TELECOMMUNICATION 
  S.W.I.F.T.
  
  Name 
  & address : SOCIETY FOR WORLDWIDE INTERBANK 
  FINANCIAL
  
  of 
  issuing TELECOMMUNICATION S.W.I.F.T.
  
  organization Avenue 
  Adele,1
  
  B - 
  1310 La Hulpe
  
  
  
  Tel: 
  +32 2 6553111
  
  Fax: 
  +32 2 6553226
  
  Structure of : 1) ICD 4 
  Digits
  
  code 
  Organization code upto 11 characters
  
  Organization name upto 250 
  characters
  
  2) 
  None
  
  Display : 
  None
  
  requirements
  
  Description of : Bank's identifiers for 
  financial transfer on the
  
  organizations S.W.I.F.T. 
  Network.
  
  covered 
  by
  
  coding 
  system
  
  Notes 
  on use : To be used for assignment of object 
  identifiers
  
  of 
  code (ISO 8824/8825)
  
  Sponsoring : 
  S.W.I.F.T.
  
  authority
  
  Date 
  of issue : September 1990
  
  of 
  ICD
  
  Additional : 
  None
  
  Comments
  
  
  
  Unfortunately an explicit abbreviated 
  name entry is not provided, although S.W.I.F.T is apparently 
  intended.
  
  
  
  This 
  would mean that the method (approved by the old CPPA TC) currently 
   yields within the approved errata for version 2, 
  either
  
  
  
  urn:oasis:names:tc:ebxml-cppa:partyid-type:S.W.I.F.T:0021
  
  
  
  or 
  
  
  
  
  urn:oasis:names:tc:ebxml-cppa:partyid-type:SocietyForWorldwideInterbankFinancialTelecommunicationS.W.I.F.T.:0021
  
  
  
  I like the shorter one. Both of these 
  are the party-id type values. Explicit values for 
  banks would presumably be defined by SWIFT and these would be the actual party 
  id values, as your subject line shows.
  
                                                                 
  
  
  Also, I should mention that, for version 
  3,   I must check whether the above two options are the only options 
  -- with Pim van der Eijk, as it seems to me that an additional shortening of 
  uuid has been under discussion for version 3. It apparently has not made its 
  way into the current working draft for version 
  3.
  
  
  
  
  PartyId 
  Element
  The PartyID element provides an 
  identifier that SHALL be used to logically identify the 
  Party. The 
  value of the PartyID element is an non-empty string that provides an 
  identifier. This is the convention used, for example, within a CPP or CPA when 
  leveraged with ebMS.
  The PartyId 
  element has 
  a single attribute: type 
  that has an anyURI [XMLSCHEMA-2] value. In a CPP 
  or CPA, the type value of the type attribute SHOULD be a URN [RFC2141] that 
  defines a namespace for the value or content of the 
  PartyId 
  element. Typically, the URN would be registered in a well-known directory of 
  organization identifiers.
  For a CPP or CPA, 
  if the type 
  attribute is not present, the value or 
  content of the PartyId element 
  MUST be a URI that conforms to [RFC2396].
  While an identifier may be understood by 
  both Parties in a CPA, messaging protocols such as ebMS recommend the use of 
  URIs. Within ebMS, the value or content of the PartyId 
  MUST be a 
  URI that conforms to [RFC2396].
  Additional 
  PartyId 
  elements MAY be present under the same PartyInfo 
  element so 
  as to provide for alternative logical identifiers for the 
  Party. In a 
  CPP, if the Party has 
  preferences as to which logical identifier is used, the 
  PartyId 
  elements 
  SHOULD be listed in order of preference starting with the most-preferred 
  identifier.
  In a 
  CPP that 
  contains multiple PartyInfo 
  elements, different PartyInfo 
  elements MAY contain PartyId 
  elements 
  that define different logical identifiers. This permits a large organization, 
  for example, to have different identifiers for different business or other 
  technical purposes.
  If 
  multiple PartyId elements occur under the same PartyInfo 
  element in the CPA, all of those PartyInfo 
  elements MUST be included in the Message 
  Header. This applies, for example, when protocol bindings such as ebMS v2 are 
  used.
  The value of the 
  PartyId element 
  is any string that provides a unique identifier. The identifier MAY be 
  understood by both Parties to a 
  CPA. 
  Several schemes exist for 
  naming identifier domains. [ISO6523] enumerates values for many domains: the 
  domain is identified by a four-digit numeric sequence. For example, the ISO 
  6523 ICD 0060 identifies the Dun & Bradstreet domain. 
  Typically, the identifier would be listed in well-known 
  directories such as DUNS (Dun and Bradstreet) or in any naming system 
  specified by [ISO6523].
  Alternatively, the code list for ISO 
  9735 (UN/EDIFACT syntax) D.3 Service simple data element 0007 (routing 
  Identification Code qualifier) can be used to name identifier domains 
  [ISO9735]. ANSI ASC [X12] Data Element I05 (Interchange ID Qualifier) element, 
  used in the ISA Interchange Control Header segment, also provides a list of 
  code values for naming domains.
  These three serve as 
  "catalogues" of schemes for naming identifier domains; processes exist by 
  which additional domains can be identified through the Registration 
  Authorities (RA):
  
    the British Standards Institute 
    (BSI): ISO 6523 
    
    ISO/TC154-UN/CEFACT Joint Syntax 
    Working Group (JSWG): [ISO9735] D.E Service simple data element 
    0007 
    ANSI ASC X12: X12 Data Element 
    I05 
  
  The following 
  example shows how URNs are used for the type attribute and for the PartyId element 
  value.
  <PartyInfo 
  partyName="CompanyA" 
  defaultMshChannelId="asyncChannelA1"
  defaultMshPackageId="CompanyA_MshSignalPackage">
  <PartyId 
  type="urn:oasis:names:tc:ebxml-cppa:partyid-type:duns">123456789</PartyId>...
  
  <tp:PartyId>urn:icann:example.com</tp:PartyId>
  The first 
  example indicates the Party's DUNS number for the organization using a type 
  attribute with a URN value. The value of the PartyId element itself is the DUNS number of 
  the organization, which is a string of digits assigned by the agency. The 
  second example shows an arbitrary URN as a PartyId 
  value. No type is indicated, but the value 
  might be a URN that the Party has registered with the Internet Assigned Numbers Authority [IANA] (http://www.iana.org) 
  to identify itself directly.
  When these naming 
  conventions are used, values are generated for the type attribute from 
  information items using a well-known directory of organization identifiers 
  such as ISO6523. A method to standardize URNs used as 
  PartyId@type values using ISO6523 is 
  described. This method adheres to the rules, requirements and recommendations 
  of a CPP or CPA, or ebMS.
  An 
  implementation SHOULD provide support, and be able to accommodate, the usage 
  of the type 
  values standardized herein (See four bullets that 
  follow).
  
    If an abbreviated name is 
    described in the item titled “Name of Coding System” within the ICD list, a 
    type attribute can be constructed by prepending: 
    “urn:oasis:names:tc:ebxml-cppa:partyid-type:” to the abbreviated name and 
    appending a colon “:” followed by the ICD value. For example, using the 
    abbreviated name D-U-N-S Number: 
    
  Abbreviated Name: “D-U-N-S 
  Number”
  Upper-camel-case resultant string: 
  “D-U-N-SNumber”
  tp:type=" 
  urn:oasis:names:tc:ebxml-cppa:partyid-type:D-U-N-SNumber:0060"
  
  Note: “0060” is the 
  ICD value of D-U-N-S Number.
  To be consistent 
  with v2 CPP/A, the value that follows remains a valid type attribute value or 
  content for the PartyId element: 
  “urn:oasis:names:tc:ebxml-cppa:partyid-type:duns”.
  
    Because an abbreviated name may 
    be omitted from the ICD list, the type attribute can always contain the 
    string derived from “Name of Coding System” expressed in upper-camel-case. A 
    value can always be constructed by pre-pending 
    “urn:oasis:names:tc:ebxml-cppa:partyid-type:” to the upper-camel-case name 
    and appending a colon “:” followed by the ICD value. For example, using the 
    formal name of the Name of Coding System: “Data Universal Numbering 
    System”: 
  
  Transformed 
  Camel-case: 
  “DataUniversalNumberingSystem”
  tp:type=" 
  urn:oasis:names:tc:ebxml-cppa:partyid-type:DataUniversalNumberingSystem:0060"
  
    Punctuation marks in these 
    formal names (such as, “/”, “-“ or “’” ) should be included unless they are 
    not allowed in URNs [RFC2141]. If the punctuation characters are not allowed 
    in URNs, then the hexadecimal escaping convention explained in [RFC2141] 
    should be followed for characters not allowed in URNs. However, spaces are 
    not allowed in URNs and should be consumed during the production of an 
    upper-camel-case string, rather than preserved in an escaped form. Words in 
    names that are all upper-case should remain so when converted to an 
    upper-camel-case string. 
    
    The ICD value should be appended 
    as the last field of the URN so that any collision between formal or 
    abbreviated names is avoided. 
    
  
  
  
  
  
  
  
  
  From:  [mailto:]
Sent: Wednesday, 
  August 05, 2009 7:38 AM
To: 
  ;
Subject: 
  [ebxml-dev] <PartyId type="?SWIFT 
  BIC?">BBBBCCLL123<\PartyId>
  
  
  
  Hi,
  
  The 
  link at http://www.oasis-open.org/committees/ebxml-cppa/documents/PartyID_Types.shtml referenced in the 
  ebcpp-2[1].1-april-5-2005-draft.doc document appears to be 
  broken.
  
  I've searched and searched, and sorry if 
  these questions have already been asked ... does anybody know where the list 
  of ebMS PartyId type URNs are maintained? Does anybody know the URN for a 
  SWIFT BIC (ISO-9362)? I would guess "urn:swift" would be insufficient and it 
  is not clear whether the URN is ISO or SWIFT. Also, ISO 6523 only gives the 
  code "0021" and not the URN. RFC 3615 doesn't help 
  either.
  
  Thanks.
  
  Charles