Re: [pcp] WGLC: draft-ietf-pcp-server-selection-04.txt comments due by AUG 18

[<[e-mail]>]

Re-,

Ooops, the proposed change in Section 4 as implemented in my local copy is:

NEW:
   As specified in Section 11.2 of [RFC6887], the PCP client must use
   different Mapping Nonces for each PCP server it communicates with.

Cheers,
Med

>-----Message d'origine-----
>De : pcp [mailto:[e-mail]] De la part de
>[e-mail]
>Envoyé : lundi 18 août 2014 14:59
>À : Simon Perreault
>Cc : [e-mail]
>Objet : Re: [pcp] WGLC: draft-ietf-pcp-server-selection-04.txt comments due
>by AUG 18
>
>Simon,
>
>My intent was not reject your proposal, but to explain the rationale for
>the current text.
>
>Below the changes I added to my local copy:
>
>Section 3:
>
>OLD:
>   For efficiency, the PCP client SHOULD use the same Mapping Nonce for
>   requests sent to all PCP server IP addresses.
>NEW:
>   For efficiency, the PCP client SHOULD use the same Mapping Nonce for
>   requests sent to all IP addresses belonging to a single PCP server.
>
>Section 4:
>
>NEW:
>   As specified in Section 11.2 of [RFC6887], different Mapping Nonces
>   must be used for requests sent to different PCP servers.
>
>Please let me know if this works for you.
>
>Cheers,
>Med
>
>>-----Message d'origine-----
>>De : Simon Perreault [mailto:[e-mail]]
>>Envoyé : lundi 18 août 2014 14:26
>>À : BOUCADAIR Mohamed IMT/OLN
>>Cc : Dave Thaler; [e-mail]
>>Objet : Re: [pcp] WGLC: draft-ietf-pcp-server-selection-04.txt comments
>due
>>by AUG 18
>>
>>Med,
>>
>>All you wrote below is true. However, the fact is I was confused for a
>>bit when I read this part, and I think I know PCP quite well. So I'm
>>pretty sure that if you give this to a random programmer there's a
>>good chance the confusion will create bugs. Better to be redundant and
>>clear IMHO. Since I'm not proposing any change to the meaning of the
>>text, I'm a bit surprised by the rejection.
>>
>>Simon
>>
>>On Mon, Aug 18, 2014 at 3:36 AM,  <[e-mail]> wrote:
>>>>   For efficiency, the PCP client SHOULD use the same Mapping Nonce for
>>>>   requests sent to all PCP server IP addresses.
>>>>
>>>>I suggest rewording for clarity:
>>>>
>>>>   For efficiency, the PCP client SHOULD use the same Mapping Nonce for
>>>>   requests sent to all IP addresses belonging to a single PCP server.
>>>
>>> [Med] The text you quoted is under "3. IP Address Selection: PCP Server
>>with Multiple IP Addresses". The text should be interpreted in that
>>context. The text currently in the draft is OK IMHO.
>>>
>>>>Different Mapping Nonces MUST still be used for requests sent to
>>>>different PCP servers.
>>>
>>> [Med] This is already in RFC6887, section 11.2. This document does not
>>modify that behavior. If you really think a note to remind that behavior
>is
>>needed, then a sentence can be added to Section 4. Thanks.
>_______________________________________________
>pcp mailing list
>[e-mail]
>https://www.ietf.org/mailman/listinfo/pcp