Re: [secdir] SECDIR review of draft-mahesh-etsi-urn-03

[Mahesh Jethanandani <[e-mail]>]

Hi Chris,

> On Jul 23, 2018, at 6:02 PM, Chris Lonvick <[e-mail]> wrote:
> 
> Hi,
> 
> I have reviewed this document as part of the security directorate's ongoing effort to review all IETF documents being processed by the IESG. These comments were written primarily for the benefit of the security area directors. Document editors and WG chairs should treat these comments just like any other last call comments. 
> 
> The summary of the review is Ready with Issues.
> 
> I am not that familiar with creating and maintaining URNs so please set me straight and move on if I'm off base here. 
> 
> Section 2, "URN Specification for ETSI" contains the following:
> 
> "ETSI will manage resource classes using the "etsi" NID and will be the authority for managing resources and associated subsequent strings. ETSI will guarantee the uniqueness of the strings themselves, or it may permit secondary responsibility for certain defined resources." 
> 
> But then says:
> 
> "ETSI may allow for use of experimental type values for testing purposes only. Note that using experimental types may create collision as multiple users may use the same values for different resources and specific strings."
> 
> It looks like RFC 8141 gives guidance that experimental use of URN namespaces is to be done using RFC 6963 (URNs for "Examples".) RFC 8184 does not address establishing or using namespaces under the subsequent strings for experimental use, but I could see that the registered owner of the namespace could make a designation for that purpose. That being said, I think that the two statements above are in conflict in the document and should be clarified.

Not necessarily.

RFC6963 says that there are three types of namespaces, formal, informal and experimental. What this draft is requesting for is the formal namespace. The RFC also says that experimental namespaces are by definition unregistered. That there is no issuing authority for experimental URNs. The owner of the NID therefore is not required to keep track of experimental namespaces. 

> 
> Perhaps it would be better to reconstruct the second paragraph to say something like:
> 
> "ETSI may allow for use of experimental type values for testing purposes only. Some experimentation may be controlled by ETSI within a subsequent string to ensure that there will be no namespace collisions among participants. Other experimentation may be controlled within a secondary namespace that may allow collisions, but this experimental use will be constrained. All other experimentation must follow the guidance set forth in RFC 6963.”

How about this?

OLD:

ETSI may allow for use of experimental type values for testing purposes only. Note that using experimental types may create collision as multiple users may use the same values for different resources and specific strings.

NEW:

ETSI may allow for use of experimental type values for testing purposes only. Note that using experimental types may create collision as multiple users may use the same values for different resources and specific strings. All experimentation must follow the guidance set forth in RFC 6963.

> 
> Just as a nit, the Security Considerations section should be revised as it seems to be a bit unclear.
> Current:
>    There are no additional security considerations other than those
>    described above, and are normally associated with the use and
>    resolution of URNs in general, which are described in Function
>    Requirements for URN [RFC1737 <https://tools.ietf.org/html/rfc1737>], Uniform Resource Names (URNs)
>    [RFC8141 <https://tools.ietf.org/html/rfc8141>].
> 
> Suggested:
>    There are no additional security considerations other than those
>    described above, and those normally associated with the use and
>    resolution of URNs in general. These are generally described in Functional
>    Requirements for Uniform Resource Names [RFC1737 <https://tools.ietf.org/html/rfc1737>], and Uniform Resource Names (URNs)
>    [RFC8141 <https://tools.ietf.org/html/rfc8141>].

Ok.

Thanks.

> 
> Best regards, Chris

Mahesh Jethanandani
[e-mail]