Re: [Pearg] Research Group Last Call for "A Survey of Worldwide Censorship Techniques"

["S. Moonesamy" <[e-mail]>]

Dear Joe,
At 03:05 PM 23-07-2020, Joseph Lorenzo Hall wrote:
>Hi, sorry for not following up sooner (going through things making 
>slides for next Monday!).
>
>I'm unclear as to what you'd like the draft to say differently here. 
>Happy to work on making it more clear but I think what you've 
>outlined is what I understand the ability for an RIR to be compelled 
>to sign routes that misdirect traffic. best, Joe

I provided an opinion of the technical aspects based on my 
understanding of how RPKI and TLS works.  RFC 8446 describes the 
properties of the secure channel defined in that document and states 
that "These properties should be true even in the face of an attacker 
who has complete control of the network ..."  My reading of the text 
in Section 3 is that those properties (Authentication, 
Confidentiality, Integrity) would no longer be true if a route was 
incorrectly "signed".

I would suggest verifying whether my interpretation (please see 
above) is correct or not.

Regards,
S. Moonesamy