Re: [Gen-art] Gen-ART review of draft-ietf-v6ops-siit-eam-01

[Tore Anderson <[e-mail]>]

Hello Dan,

Thank you very much for your review! My comments are in-line.

* "Romascanu, Dan (Dan)" <[e-mail]>

> 1.       In the second paragraph of the Introduction I suggest s/The
> Explicit Address Mapping Table does not replace/Translation using the
> Explicit Address Mapping Table does not replace/

Implemented.

> 2.       In section 2 I would suggest s/doing so may result in a new
> set of undesired properties/doing so may result in a new set of
> undesired consequences/

Implemented.

> 3.       Section 3.2:
> 
>    When translating a packet between IPv4 and IPv6, an SIIT
>    implementation MUST individually translate each IP address it
>    encounters in the packet's IP headers (including any IP headers
>    contained within ICMP errors) according to Section 3.3.  See
>    Section 4 for certain exceptions to this rule.
> 
>        As we are talking about exceptions to the rule, is not SHOULD
> more appropriate than MUST?

Well, the point I am trying to get across is that it is a indeed MUST,
unless the specific condition in Section 4 is present - but *only* then.

I believe that simply replacing "SHOULD" with "MUST" would result in a
too loose requirement, as that would allow from deviation from Section
3.3 even in cases other than those described in Section 4.

I'm struggling a bit to come up with a decently worded sentence that
captures this point better. Maybe this?

    When translating a packet between IPv4 and IPv6, an SIIT
    implementation MUST individually translate each IP address it
    encounters in the packet's IP headers (including any IP headers
    contained within ICMP errors) according to Section 3.3, except for
    any address for which Section 4 explicitly states that the EAM
    algorithm MUST NOT be used.

If you have any better suggestions, I am all ears.

> 4.       Sections 4.2.1 and 4.2.2 present two alternative approaches
> for hairpinning support. Yet, the opening sentence in 4.2.1. a
> keyworded MUST, while the opening sentence in 4.2.2. does not:
> 
> 
> 
>    When the simple hairpinning feature is enabled, the translator MUST
> 
>    behave according to the following rules when translating from IPv4
> to
> 
>    IPv6:
> 
> 
> 
>    When the intrinsic hairpinning feature is enabled, the translator
> 
>    behaves as follows when receiving an IPv6 packet:
> 
> 
> 
>    It seems that either MUST is to be used in both, either in none.

Implemented.

You can see the changes in diff format here:

https://github.com/toreanderson/ietf/commit/68c37731e7b1f7024dcd1c19b491b1b4f3ff963f

Please have a look, and let me know if further changes are necessary.
Thank you again!

Best regards,
Tore Anderson