Re: default-iids: dropping requirement 1 in Section 3

[Fernando Gont <[e-mail]>]

On 05/20/2016 05:06 PM, Alissa Cooper wrote:
> 
>> On May 18, 2016, at 1:26 AM, Alissa Cooper <[e-mail]> wrote:
>>
>> What if we dropped requirement #1 in Section 3, but left the rest the same? This would open the door such that people who wanted to comply with the SHOULD-level recommendation in this spec in the presence of randomized MAC addresses still could (e.g., by implementing 7217 and using the randomized MAC as the Net_Iface parameter), but directly embedding a MAC address in the IID would still go against the SHOULD NOT recommendation here.
> 
> I don’t think I saw anyone respond to this suggestion, so I’d like to ask again for people’s opinions about this.
> 
> To my eye this is a nice compromise.

That's a bad idea.

The reason why is  draft-gont-predictable-numeric-ids-00, RFC4941, and
even some RFC (?) by the IAB.

The root of this whole problem is embedding a layer-2 id in a layer-3
field in the first place.

Thanks,
-- 
Fernando Gont
SI6 Networks
e-mail: [e-mail]
PGP Fingerprint: 6666 31C6 D484 63B2 8FB1 E3C4 AE25 0D55 1D4E 7492