Re: [Hipsec] Minor correction to rfc7401 4.4.4 state machine

[Gonzalo Camarillo <[e-mail]>]

Hi Tom,

Yes, the trend is to only "verify" errata that can clearly lead to interop problems. Other errata are typically "held for document update".

Cheers,

Gonzalo

Sent from my mobile


---- Tom Henderson wrote ----

Gonzalo,
I had a look at the criteria here:

https://www.ietf.org/iesg/statement/errata-processing.html

and it states that this condition should be satisfied to log an erratum:

 Only errors that could cause implementation or deployment problems or significant confusion should be Verified.

I'm wondering about whether this meets the criteria.  On the one hand, this is a mainstream state transition (I1-SENT gets to I2-SENT upon receiving an R1, not an R2).  On the other hand, this section is clearly informational (section 4.4) and the transition is correctly noted in the Table 3 in this section (also informative) and in the normative Section 6.8.

So, I wonder whether this is in the category of "could cause significant confusion" or rather whether an implementer would just recognize this as an obvious typo in conflict with other parts of the specification.

Any opinions about this?

- Tom

On 06/29/2015 01:36 PM, Gonzalo Camarillo wrote:> Yes, please log an erratum.
>
> Thanks,
>
> Gonzalo
>
> On 29/06/2015 11:20 PM, Tom Henderson wrote:
>>
>>
>> On 06/28/2015 09:51 AM, Darren Lissimore wrote:> Hey all;
>>>
>>> Read through 7401, got to section 4.4.4
>>> ​page 36.
>>> There's an error in the transition from I1-SENT to I2-SENT
>>> ​.​
>>> The draft has it as  recv R2, send I2 and
>>> ​it ​
>>> should be
>>> ​probably be ​
>>> recv R1, send I2
>>> ​as per Table 3 page 29   "recevie r1, process" trigger.
>>>
>>> ​Darren Lissimore