ivmarkov

> @ivmarkov thinking about it again, we could remove the v4.4 option in the advanced path now. What do you think? Actually, I'm fine either way. Nobody is supposed to...

Hmmm, or shall we just clear the FailSafe when we switch from BLE to Wifi? That should also be app-induced though? And we'll lose the NOC state...

> > Hmmm, or shall we just clear the FailSafe when we switch from BLE to Wifi? That should also be app-induced though? And we'll lose the NOC state... >...

> Once an `AddNOC` is received, this should update the mode to `SessionMode::Case(new_fabric_index)` thus fixing the problem in the current scenario? Yes, but **only if** after `AddNOC` the comissioner does...

In any case, let me try what you suggest.

@kedars Finally back to this issue today, after a flurry of other bugfixes. The bad news: no, your suggestion won't work :( ... and yes, we DO receive the second...

Here, a shortened log for your entertainment (all over BLE btw). Grep the `!!!!!!!!!!` stuff I've embedded: ``` >>>>> BTP 4E:19:47:80:FA:3A [S,SID:0,CTR:a8ed7fe,SRC:6da43a0e9c970244][I,EID:646a,PROTO:0,OP:20] SC::PBKDFParamRequest !!!!!!!!!! ^^^ Plain text session starts (SID:0),...

I think this is now ready for review. The reason why this PR has a bigger changeset is because I changed the representation of `fabric_index` from `u8` to `NonZeroU8` in...

Hm, maybe [this one](https://github.com/project-chip/connectedhomeip/tree/master/src/python_testing)?

Adressed with #289 and #290