Re: Connected Cars. Upgradable/Replaceable IoT systems. Re: Managing Long-Lived CA certs
Peter Gutmann <[email protected]>
| Newsgroups | gmane.ietf.x509 |
|---|---|
| Message-ID | <[email protected]> |
Robert Moskowitz <[email protected]> writes: >The IEEE 1609.2 standard for Vehicle safety messaging has a 'monster' PKI >with certificate management. To put it mildly. Not just a monster PKI, a monster in general. They invented their own gratuitously incompatible way of doing everything possible (message security, certificates, you name it, including a pile of novel security mechanisms never before deployed at any real-world scale). The thing with X.509, CMS, PGP, whatever you want to use is that after about twenty years of public naming and shaming vendors have got at least some of the bits right (but see for example the thread currently running on mozilla- dev about CAs all over the world issuing certs for domain names that can't be validated, in other words that were never checked by the CA before the certs were issued). OTOH the 1609.2 stuff, which goes way beyond what any standard public CA has ever attempted (e.g. SCMS or Secure Credential Management System, which is... no, it's too horrible to go into) will be used in a closed, non-public environment where little if anything will ever be tested or checked for correctness. Until the Black Hat and Defcon presentations start appearing... >I worked on three telematics certificate systems. I got exposed to 1609.2. My response was that there simply wasn't enough money in existence to get me to try and make that thing work. Peter. _______________________________________________ pkix mailing list [email protected] https://www.ietf.org/mailman/listinfo/pkix