Technically unpredictable is probably a better term, but random implies that. I'll leave it to the authors to determine the best language. (I'm ok with what the 04 draft now says.) - Bernie -----Original Message----- From: Fernando Gont [mailto:fgont@xxxxxxxxxxxxxxx] Sent: Tuesday, February 16, 2016 4:39 AM To: Bernie Volz (volz) <volz@xxxxxxxxx>; Robert Sparks <rjsparks@xxxxxxxxxxx>; Tomek Mrugalski <tomasz.mrugalski@xxxxxxxxx>; General Area Review Team <gen-art@xxxxxxxx>; ietf@xxxxxxxx; dhcwg@xxxxxxxx; draft-ietf-dhc-dhcpv6-privacy.all@xxxxxxxx Subject: Re: [Gen-art] Gen-ART LC review: draft-ietf-dhc-dhcpv6-privacy-03 Bernie, On 02/15/2016 06:37 PM, Bernie Volz (volz) wrote: > Perhaps we should get away from whether something is easy or difficult > to implement or whether the algorithm may be more (or less) efficient. > > I think the point of this material is to ENCOURAGE random assignment > rather than sequential to improve privacy- so keep it at that. Let > implementers worry about how efficient an algorithm is? There are a number of operational and security/privacy considerations that depend on the actual algorithm that you employ. I'm not sure what you mean by "random". IID=random() everytime an address has to be leased? -- or do you really mean "unpredictable"?. At the end of the day, what you want is to specify desired properties. Besides, assignment of sequential addresses should be banned. They have always been a bad idea (see <https://tools.ietf.org/html/draft-gont-predictable-protocol-ids>) and they shouldn't even be considered an option. Thanks, -- Fernando Gont SI6 Networks e-mail: fgont@xxxxxxxxxxxxxxx PGP Fingerprint: 6666 31C6 D484 63B2 8FB1 E3C4 AE25 0D55 1D4E 7492