On Thu, Sep 29, 2011 at 10:09 AM, Roland Bless <roland.bl...@kit.edu> wrote: > Hi, > > On 29.09.2011 15:44, Christopher Morrow wrote: >> have to help in the educational process a bit, but hiding behind >> 'private addressing' and 'we never want to ... oops, we connected to >> the internet!' just isn't working today. > > As a general statement fine, but in our use case you > a) need stable addresses which can be used independently > of any other existing connectivity (whether ULAs > or global PI addresses doesn't matter so much)
Sure, a structure to build that upon 'reboot' seems like a simple task to accomplish... lots of things do this already, it's not new or scary, it does take some planning and thought early in the process though. > b) explicitly consider cases where you get connected > to the Internet (w/o putting safety at risk) right, each part (addressable item in the total system) needs to understand it's place in the security world, protect itself and it's neighbors. It's fine to have (and probably a great thing) some security measures at the 'borders' (gateways between "internal" and "external") but those can not be the sole guardians. Quality code, quality security architecture, quality interfaces between components (hardware and software) are all key to making the vehicular environment safe. my larger point wasn't really about your case though... -chris -------------------------------------------------------------------- IETF IPv6 working group mailing list ipv6@ietf.org Administrative Requests: https://www.ietf.org/mailman/listinfo/ipv6 --------------------------------------------------------------------
