Menu

#53 R0886 uses different URI/IRI standard

D4r1
closed
None
False
not added
D4
886
2022-06-27
2020-09-16
No

Comment

It uses RFC 3986, but it should be RFC 2396 or RFC 3987.

Rationale / Proposed change

http://www.w3.org/TR/xmlschema-2/#anyURI references RFC 2396 instead of RFC 3986.
http://www.w3.org/TR/xmlschema11-2/#anyURI references RFC 3987 instead of RFC 3986.

Related

Closed: P11073-10700 Comments: #94

Discussion

  • Björn Andersen

    Björn Andersen - 2020-10-12
    • status: unread --> open
    • assigned_to: Björn Andersen
     
  • David

    David - 2020-11-06
    • status: open --> closed
     
  • David

    David - 2020-11-06

    It looks like the newer RFC does only impose restrictions; 3986 should be backwards compatible such that a 2396 parser can always parse 3986 URIs. Requirement has been relaxed to SHOULD.

     
    • Maximilian Pilz

      Maximilian Pilz - 2020-11-13

      I do not see this parser statement being true, for example "http:" is valid according to RFC 3986 and can be parsed by RFC 3986 parsers, but it is not allowed in RFC 2396 and thus may be rejected by an RFC 2396 parser.

      This is also stated in RFC 3986:

      "The ABNF has been corrected to allow the path component to be empty.
      This also allows an absolute-URI to consist of nothing after the
      "scheme:", as is present in practice with the "dav:" namespace
      [RFC2518] and with the "about:" scheme used internally by many WWW
      browser implementations."

      Though your argumentation is wrong, it really is a matter of the corresponding XML schema being used or not, because it enforces RFC 2396 and thus would eliminate the possibility of such URI.

      Then the question arrises, if those URIs that are valid for both are enough for the use case and if different segment names are a problem for interoperability.

       

      Last edit: Maximilian Pilz 2020-11-13
      • Björn Andersen

        Björn Andersen - 2020-11-20

        Different segments names could lead to different parsing results!
        Is this a potential problem?

         
        • Maximilian Pilz

          Maximilian Pilz - 2021-06-07

          Yes, it is a potential problem.

           
  • Björn Andersen

    Björn Andersen - 2020-11-20
    • status: closed --> open
     
  • Björn Andersen

    Björn Andersen - 2020-11-20
    • assigned_to: Björn Andersen --> David
     
  • David

    David - 2020-12-01
    • Draft: -->
    • Milestone: --> D1r1
     
  • David

    David - 2021-02-22

    We shouldn't rely on outdated RFCs. For this reason, keep using RFC 3986 and fix the URI standard reference in BICEPS, see [ieee11073-10207:#209]

     
    👍
    1

    Related

    11073-10207 Revision: #209

  • David

    David - 2021-02-22
    • status: open --> wont-fix
    • Draft: --> D1
     
  • Maximilian Pilz

    Maximilian Pilz - 2021-06-07
    • Description has changed:

    Diff:

    --- old
    +++ new
    @@ -1,7 +1,8 @@
     **Comment**
    
    -It uses RFC 3986, but has it has to be RFC 2396.
    +It uses RFC 3986, but it should be RFC 2396 or RFC 3987.
    
     **Rationale / Proposed change**
    
     http://www.w3.org/TR/xmlschema-2/#anyURI references RFC 2396 instead of RFC 3986.
    +http://www.w3.org/TR/xmlschema11-2/#anyURI references RFC 3987 instead of RFC 3986.
    
     
  • Maximilian Pilz

    Maximilian Pilz - 2021-06-07
    • status: wont-fix --> open
     
  • Maximilian Pilz

    Maximilian Pilz - 2021-06-07
    • Milestone: D1r1 --> D3
     
  • Maximilian Pilz

    Maximilian Pilz - 2021-06-07
    • summary: R0886 uses wrong URI standard --> R0886 uses wrong URI/IRI standard
     
  • Maximilian Pilz

    Maximilian Pilz - 2021-06-07
    • summary: R0886 uses wrong URI/IRI standard --> R0886 uses different URI/IRI standard
     
  • David

    David - 2021-06-10
    • Milestone: D3 --> D3r1
     
  • Björn Andersen

    Björn Andersen - 2021-07-13
    • Ballot comment: --> not added
     
  • Björn Andersen

    Björn Andersen - 2021-07-16
    • status: open --> wont-fix
     
  • Björn Andersen

    Björn Andersen - 2022-03-10
    • status: wont-fix --> open
    • Must be satisfied: --> False
    • Draft: D1 --> D4
    • Milestone: D3r1 --> D4r1
     
  • Björn Andersen

    Björn Andersen - 2022-03-10

    @d-gregorczyk argues against this solution here: https://sourceforge.net/p/opensdc/ieee11073-20701/47/#81c7

    If we do not want to require XML Schema 1.1, we may need to introduce a dirty fix like this:

    An SDC Participant SHALL limit values of pm:InstanceIdentifier/@Root to those matching the form scheme ":" hier-part as defined in [RFC 3986] Clause 3 and scheme ":" hier_part as defined in [RFC 2396] Clause 3 except for "biceps.uri.unk" and "sdc.ctxt.loc.detail".

     

    Last edit: Björn Andersen 2022-03-10
  • Björn Andersen

    Björn Andersen - 2022-06-27
    • status: open --> closed
    • assigned_to: David --> Björn Andersen
     
  • Björn Andersen

    Björn Andersen - 2022-06-27

    Accepted as proposed on 2022/03/10.

     

Log in to post a comment.

MongoDB Logo MongoDB