Re: [93attendees] Reminder: CITIZENFOUR screening *tomorrow* at 7pm

[Christian Huitema <[e-mail]>]

On Thursday, July 30, 2015 6:55 AM, [e-mail] wrote:

> Snowden entered in his answers very specifically on the work of the IETF and warned 
> about the recently proposed idea to provide middle boxes through a new Protocol with 
> the working title "Spud" with some data, to prevent disconnections or deep packet 
> inspection. His plea for the restoration of end to-end principle - "the easiest path is the best" - 
> was applauded extensively by many IETF participants. 

Yes, that's what we heard. The Register's reporter got very confused, but the German transcript is correct.

> Snowden hoped for the reduction of  metadata by other IETF work, e.g. the effort towards a 
> more privacy friendly DNS including the developments of DANE and DNSSEC. 

Both DANE and DNSSEC can improve security, and help making sure that encryption works end-to-end. But the specific issue addressed here is the leakage of meta-data in DNS queries, which is addressed by the QName minimization proposed in DNSOPS and by the proposed work for stub-to-recursive-resolver encryption in DPRIVE.

This was an answer to a question by Juan Carlos, and the general idea is that we have to be concerned not just with encryption of content, but also with the leakage of meta-data at various layers in the stack. Snowden was for example commenting on listening devices that record radio traffic. Imagine drones circling abve cities and recording the IMSI of cell phones, or the MAC addresses used by the clients of Wi-Fi networks. That can lead to continuous tracking of devices and their owners -- and we should be doing something about it.

-- Christian Huitema