Re: RA "requires" DHCPv6 ?

On Sat, 2012-03-31 at 21:43 +0200, Fernando Gont wrote: > > Or you could use one's statically assigned address. > > Oh, yeah. And you could have opted to not send the RS in the first > place... and what? :-) Um, I'm not sure why we're arguing. I'll still see unsolicited RAs at regular intervals,

New revision of draft-gont-6man-stable-privacy-addresses

Folks, I've just posted a revision of the aforementioned I-D. It is available at: It contains a number of improvements and clarifications, based on the questions received during the 6man wg meeting, but also thanks to a nu

Re: RA "requires" DHCPv6 ?

On 03/31/2012 03:12 PM, Karl Auer wrote: > On Sat, 2012-03-31 at 14:10 +0200, Fernando Gont wrote: >>> Anyway, I've been working on the basis that the M and O flags are >>> advisory and not prescriptive. That is, they do not *require* the >>> host to do anything. >> >> Exactly: They do not REQUIRE

Re: RA "requires" DHCPv6 ?

On Sat, 2012-03-31 at 14:10 +0200, Fernando Gont wrote: > > Anyway, I've been working on the basis that the M and O flags are > > advisory and not prescriptive. That is, they do not *require* the > > host to do anything. > > Exactly: They do not REQUIRE you to do DHCPv6. You MAY want not to do >

Re: [6man] Stable privacy addresses (upcoming rev)

On 03/31/2012 07:32 AM, Christian Huitema wrote: >> If the regime controls the local-link, then as far as >> address-tracking is concerned, you're toast. -- They could sniff >> the network and log the address->MAC mappings, have RAs require you >> to do DHCPv6 and then have DHCPv6 assign you a cons

Re: RA "requires" DHCPv6 ?

Hi, Karl, On 03/31/2012 12:57 AM, Karl Auer wrote: > In a discussion titled "Stable privacy addresses (upcoming rev)", > Fernando Gont said: >> They could [...] have RAs require you to do DHCPv6 and then have >> DHCPv6 assign you a constant address, etc. > > What interests me here is the phrase "

Re: RA "requires" DHCPv6 ?

On Sat, 2012-03-31 at 23:03 +1100, Hesham Soliman wrote: > >There was a bunch of stuff about the M and O flags in RFC2462, almost > >all of which was removed in RFC4862. In RFC2462, the word > >"should" (*not* capitalised) was used, along with phrases like "is to > >be". > > => "should" does not n

Re: RA "requires" DHCPv6 ?

Just a quick comment below >There was a bunch of stuff about the M and O flags in RFC2462, almost >all of which was removed in RFC4862. In RFC2462, the word >"should" (*not* capitalised) was used, along with phrases like "is to >be". => "should" does not need to be capitalised to indicate that it