Re: [6tisch] MSF Shepherd review

[Tengfei Chang <[e-mail]>]

Hi Yatch,

The session is distinguished by the parent, there is no such case that two
MSF sessions have a 'common parent'.
MSF session to parent is 1 to 1 mapping relationship.
MSF session should be ends as long as a neighbor is un-selected as parent.

In the text, I agree we only details how MSF deal with one parent.
I think Pascal agree with this but just don't want to limit MSF to just one
parent, which is clarified in the paragraph I wrote above.

Tengfei



On Fri, Dec 6, 2019 at 9:43 PM Yasuyuki Tanaka <[e-mail]>
wrote:

> Hi Tengfei,
>
> How do we distinguish multiple MSF sessions?
> What if two MSF sessions have a same "selected parent", and then one of
> them selects another selected parent? How many negotiated cells should
> be taken over to the new selected parent?
>
> These are not covered by the current text, and I think they are out of
> the scope of MSF.
>
> > This specification only describes how MSF works with one routing parent,
> which is phrased as "selected parent".
>
> So, yes, I believe we should have this case only in the text. If we
> mention some other possibilities without concrete ideas to implement,
> they would just confuse future readers.
>
> I think, Pascal has a different opinion.
>
> Best,
> Yatch
>


-- 
——————————————————————————————————————

Dr. Tengfei, Chang
Postdoctoral Research Engineer, Inria

www.tchang.org/
——————————————————————————————————————