[dhcwg] Comments on draft-halwasia-dhc-dhcpv6-hardware-addr-opt-01

[Tomek Mrugalski <[e-mail]>]

I do support this draft and its adoption.

I will provide more detailed comments later.

It is useful and I'm involved in more than one implementation that is
potentially interested in implementing this option.

I like the changes made between -00 and -01.

One notable thing that should be specified is the actual content of the
link-layer address used by a client. I think it should be clarified that
this is supposed to be link-layer address from the interface that was
used to contact the server.

Please note that this is slightly different than DUID-LL or DUID-LLT
link-layer. DUIDs state that client picks one ll address, generates DUID
and sticks to it. For DUIDs it does not have to be the same interface
that is used for communication.

Think about a client with 5 interfaces. It will use the same DUID over
all 5 interfaces, but will use 5 different LINK_ADDR option.

Oh, and the draft should be renamed once adopted - it is no longer
called hardware option. Something like
draft-ietf-dhc-dhcpv6-link-layer-option will do just fine.

Tomek