Re: Packet number encryption

[Stephen Farrell <[e-mail]>]

Hiya,

On 04/02/18 19:14, Fossati, Thomas (Nokia - GB/Cambridge) wrote:
> On 04/02/2018, 18:51, "Stephen Farrell" <[e-mail]>
> wrote:
>> In addition, I'd like to point out a flaw in the above - the
>> question is posed in terms of an "end user client" making decisions
>> for "end users not particularly concerned" with a privacy enhancing
>> feature. I don't believe that programmers can really determine such
>> things (no matter how often someone says they "care deeply";-)
> 
> In user agents I guess this would be handled as a configurable
> user's preference, not a as hard-coded value.  We could have it MTI
> and on by default.  Software vendors who produce user facing
> technology have a strong incentive to do this kind of things right.

Hmm, it's a bit of an aside, and likely not decidable, but I'm not
sure you're right that implementers (whether vendors or open-source
implementers or whatever) are always driven by end-user's concerns.

And in this case, I'm not sure that an application-by-application
approach even works - even if N-1 of the N applications I use do a
great job wrt linkability, when the N-th application is careless I
can lose all benefits. (And yes, combinations of TCP and DHCP can
result in the same bad outcome, but that's not a reason to do less
well than we can IMO.)

So I still hope the WG study the trade-offs and make doing the best
we can for privacy the only option.

S.



> 
> Cheers!
> 

-- 
PGP key change time for me.
New-ID 7B172BEA; old-ID 805F8DA2 expires Jan 24 2018.
NewWithOld sigs in keyservers.
Sorry if that mucks something up;-)