Re: [Gen-art] Gen-ART Last Call review of draft-ietf-manet-olsrv2-dat-metric-08

[Paul Kyzivat <[e-mail]>]

Henning,

OK. I'm now happy with -10.

	Thanks,
	Paul

On 11/26/15 7:30 AM, Henning Rogge wrote:
> Am 25.11.2015 um 16:11 schrieb Paul Kyzivat:
>> Henning,
>>
>> I'm not trying to second guess the intent. My issues are solely with the
>> presentation. And clearly this is not a subject I'm knowledgeable about,
>> so I'm going solely from what I read in this draft without any other
>> context. Somebody who has the proper context to be reading this may well
>> not find it at all confusing. Nevertheless, see inline.
>
> No problem...
>
>> On 11/25/15 2:05 AM, Henning Rogge wrote:
>>> Am 24.11.2015 um 17:32 schrieb Paul Kyzivat:
>>>> So, if I do 9.3 for every packet, and then also do 9.4 for every HELLO
>>>> message within that packet, then I could end up incrementing
>>>> L_DAT_received[TAIL] *twice*.
>>
>>> If you look at the head of section 9.3 you will notice that this section
>>> will only be processed in the presence of a packet sequence number. One
>>> of the first things 9.3 does is to set the L_DAT_last_pkt_seqno
>>> variable.
>>
>> OK. My mistake. Because 9.3 sets L_DAT_last_pkt_seqno it won't get
>> updated again in 9.4.
>>
>> But then, what is the purpose step 3 of 9.4? If 9.3 is always applied
>> before 9.4, then L_DAT_last_pkt_seqno can never be undefined in 9.4.
>
> It could be that you have a neighbor that will never deliver a packet
> sequence number, so 9.3 will never apply.
>
> Because I am not sure if packet (9.3) or message (9.4) processing is
> done first, I make sure to overwrite the data in the current element of
> the queue in 9.3 when I discover the packet sequence number.
>
> Henning Rogge
>
> P.S.: I will be away from work in December, so I added my private email
> [e-mail] to this thread.
>
>