Sky blocks Newzbin, important legal and technical questions need answering

James Firth james2 at jfirth.net
Fri Dec 16 20:36:15 GMT 2011


Ian Batten wrote:
> Why would it be a privacy nightmare in a co-lo?  And outside a co-lo,
> RFC4941 solves the problem and is the default on both OSX (10.7 and on;
> you have to enable it on 10.6) and on XP SP3 onwards.  I presume Linux
> gets it right (/proc/sys/net/ipv6/conf/all/use_tempaddr).
> 
<snip>
>
> But there's no reason for a hosting company to use blocks at all.  As
> of today, if you buy a simple hosting arrangement, you get a single IP
> number for your single host.  Why would you expect, or need, more than
> a single IP number within your co-lo's network?  

I had this argument at the IPv6 "launch party" in London (an industry bash
designed to push IPv6 adoption).  It was explained to me then, by people who
should know, that the original intention was to merge MAC and IP addresses.

Hence the rightmost 64 bits - the "interface identifier" are reserved for
the MAC address or equivalent, and IPv6 addresses are designed to be
allocated in blocks of /64.  That's why no end "customer" should ever get
less than a /64.

Now I hear you cry, since MAC addresses are in theory unique, there's no
need to give every customer a /64... But there is, really.  MAC fiddlers,
etc.

An obvious privacy nightmare standards have emerged such as RFC 3041; which
I think, Ian, might describe similar behaviour to that mentioned in your
earlier post.

James Firth






More information about the ukcrypto mailing list