Andy,

On 22/10/2014 13:00, Andy Newton wrote:
> On 10/21/14, 6:49 AM, "Hollenbeck, Scott" <[email protected]> wrote:
> 
>>> Major Issues:
>>> -------------
>>>
>>> Section 3.1.1 says:
>>>
>>>   "The restricted
>>>    rules to write a text representation of an IPv6 address [RFC5952]
>>> are
>>>    not mandatory."
>>>
>>> Why not make 5952 at least a SHOULD? Personally, I would make it a
>>> MUST. As 5952 itself states,
>>> the ambiguity of the RFC 4291 format creates many problems. 5952 in any
>>> case requires that
>>> "all implementations must accept and be able to handle any legitimate
>>> RFC 4291 format",
>>> so making conformance with 5952 a SHOULD or MUST won't break anything.
>>>
> 
> Thanks for pointing out that sentence from 5952. Given that, though I
> don¹t see a MUST has helpful and in fact it might be confusing. My
> recommendation is:
> 
> OLD
> 
>   Any valid IPv6 text address format [RFC4291] can be used, compressed or
> not compressed.  The restricted
>   rules to write a text representation of an IPv6 address [RFC5952
> <http://tools.ietf.org/html/rfc5952>] are
>   not mandatory.
> 
> NEW
> 
>   Any valid IPv6 text address format [RFC4291] can be used The rules to
> write a text representation of an
>   IPv6 address [RFC5952] are RECOMMENDED.
> 

Yes, I'm fine with that.

> 
> 
>>> Minor issues:
>>> -------------
>>>
>>> Section 3.1.1 says:
>>>
>>>   "However, the zone id [RFC4007] is not appropriate in
>>>    this context and therefore prohibited."
>>>
>>> Agreed, but you probably also need to exclude the extended URI syntax
>>> for this [RFC6874],
>>> since your base reference is the URI syntax [RFC3986] which is updated
>>> by 6874.
>>> This is just for clarification, since 6874 does not change the ABNF
>>> production
>>> for IPv6address.
> 
> 
> Do you mean that we should change the reference of IPv6address from RFC
> 3986 to RFC 6874? If so, that¹s a good suggestion and I agree. Otherwise,
> I¹m confused.

6874 slightly extends the syntax in 3986, but you want to explicitly
exclude that extension. So I think this would cover it:

OLD:

   The IPv4 and IPv6 address formats supported in this query are
   described in section 3.2.2 of [RFC3986], as IPv4address and
   IPv6address ABNF definitions.  Any valid IPv6 text address format
   [RFC4291] can be used, compressed or not compressed.  The restricted
   rules to write a text representation of an IPv6 address [RFC5952] are
   not mandatory.  However, the zone id [RFC4007] is not appropriate in
   this context and therefore prohibited.

NEW:

   The IPv4 and IPv6 address formats supported in this query are
   described in section 3.2.2 of [RFC3986], as IPv4address and
   IPv6address ABNF definitions.  Any valid IPv6 text address format
   [RFC4291] can be used, compressed or not compressed.  The
   rules to write a text representation of an IPv6 address [RFC5952] are
   RECOMMENDED.  However, the zone id [RFC4007] is not appropriate in
   this context and therefore the corresponding syntax extension
   in [RFC6874] MUST NOT be used.

(I sort of feel obliged to apologise, since there has been confusion
in this area since RFC 2732, although the root problem was the choice
of the colon in both IPv6 address syntax and URI syntax.)

    Brian

_______________________________________________
Gen-art mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/gen-art

Reply via email to