Hi Jonathan,
Sorry, in previous email I missed your comment about section 4.3.5.
Your question is: if the SF manages several slotframes, when retrieving the 
LIST, how to differentiate them, since they appear as a tuple 
(slotOffset,channelOffset). My question is if it is worthy to make the 
differentiation when receiving all the cells. Or this will be handled by the 
SF? IMO an example will be good. 
According to my understanding, SF0 uses Metadate to express slotframe ID, see 
the description of Metadata in 6P draft. Thus, it is SF0 who handles this. For 
example, SF0 can define in this way: if specific slotframe ID in Metadata, then 
the Cells in the CellList of Response belong the specific slotframe; if setting 
some value like FF in Metadata to say all of slotframes, then the CellList of 
Response will include all of cells scheduled between the two neighbors. But, 
this should be defined and explained in SF0 draft. What do you think?
Thanks Qin    

 On Wednesday, June 7, 2017 3:54 PM, Qin Wang <[email protected]> wrote:
 

 Hi Jonathan,
Thank you very much for your review. Please see inline.
Qin


    On Wednesday, June 7, 2017 11:21 AM, Jonathan Muñoz 
<[email protected]> wrote:
 

 Dear 6P authors, W.G,
I read the 6top Protocol draft - v05, and I find it very clear. I have some 
questions and few comments. Hopefully you'll find them useful.  
Best,
Jonathan Munoz
------------------------------ ------------------------------ 
-------------------------->Section 4.1: "If no dedicated cells are scheduled 
between nodes A and B, shared   cells MAY be used. -> Is there another option? 
if not, shouldn't say MUST?[Qin]: I think there is another option, i.e. do 
nothing if no dedicated cells are scheduled between node A and node B. What do 
you think?
->Section 4.3.2:     "  All cells in the CellList MUST already be scheduled 
between the two nodes and must match the CellOptions field. If node A puts 
cells in its CellList that are not already scheduled between the two nodes and 
match the CellOptions field, node B replies with a RESET return code"
   Section 4.3.3:     "Upon receiving the request, node B's SF verifies that 
all the cells in the Relocation CellList are indeed scheduled with node A, and 
are associate the options specified in the CellOptions field. If that check 
fails, node B MUST send a 6P Response to node A with return code CELLLIST_ERR."
Both errors give a different response but it looks to me that they have the 
same source -mismatch in their schedule-. Is it ok?
[Qin] You are right. We change error code in 4.3.2 to CELLLIST_ERR.
->Section 4.3.5:      if the SF manages several slotframes, when retrieving the 
LIST, how to differentiate them, since they appear as a tuple 
(slotOffset,channelOffset). My question is if it is worthy to make the 
differentiation when receiving all the cells. Or this will be handled by the 
SF? IMO an example will be good.
->Figures: 17, -6P COUNT Response and Confirmation-                19, -6P LIST 
Response and Confirmation-                21, -6P CLEAR Response and 
Confirmation-
    Will these transaction have a confirmation message? it seems to me they are 
a 2-step transactions, always.
[Qin] accepted.
-> Sequence Number: will it increase +1 despite the failure of the previous 
transaction?
[Qin]: I think so. Is there any particular reason for not doing in this way?

------------------------------ ------------------------------ 
-------------------------   

-- 
Jonathan MuñozPhD student at Gridbee Communications & INRIA Paris+33 (0) 6 59 
05 93 83_______________________________________________
6tisch mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/6tisch


   _______________________________________________
6tisch mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/6tisch


   
_______________________________________________
6tisch mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/6tisch

Reply via email to