Michael English:
> Anonym,
>
> Please Cc me with any replies. For some reason, I am unable to subscribe to the list.
Got it!
> Anonym:
>> Note that I haven't chimed in with my opinion yet. :) I think intrigeri
>> is just careful: privacy features are routinely exaggerated in their
>> efficacy and can easily have many unexpected effects; for instance,
>> playing with such parameters often introduces fingerprintability
>> ("this user seems to have enabled the non-default feature X, which
>> Tails enables, so +1 indication that this is a Tails user"). It's
>> hard to weigh weigh such advantages and disadvantages against each
>> other, especially when the domain (blockchain) is not well-known,
>> which at least is the case for me (intrigeri has read up in the past
>> weeks for other reasons, so he's definitely in a better position to
>> evaluate).
>
> Increased fingerprintability is not at all relevant here. Consider
> adding uBlock Origin to the Tor Browser as an example of increasing
> the fingerprint left by Tails users. Websites could check whether a
> client using Tor is blocking ads to narrow down that the user is a
> Tails user. The coin selection that I recommend changing happens
> entirely offline. No remote servers are involved in the creation of a
> new transaction. Once the transaction is broadcast to the Bitcoin
> network, the fingerprintability actually goes down as the transaction
> looks more generic. Remember that all transactions will have inputs
> and outputs that appear the same to the network. The change here is
> selecting inputs that reveal less about the user’s total bitcoin
> balance.
Note that "fingerprintability" was an example said in the spirit of "deviating from the default options often makes observers able to partition users" [0]. I'm happy to hear this feature in fact will defend against that somewhat. FWIW, with my extremely limited understanding of the blockchain, what you say sounds pretty convincing. :)
[0] "4. Case study: against options",
https://www.freehaven.net/anonbib/cache/usability:weis2006.pdf
>> So, if the decision comes down to me, I'd delegate to our community.
>> That's also hard. Now we have two voices in favour, none against,
>> and some probably good arguments for those that know how the
>> blockchain works. If we can get some argument why enabling this
>> feature won't have nasty consequences (e.g. increased
>> fingerprintability) and get some more people to join in support for
>> this change, I'd do it. I am afraid this is the best I can do.
>
> Is anyone else in favor or against this change?
I wouldn't count my vote, due to my blockchain ignorance. :)
> My attempt is to make privacy a default option in Tails.
Of course, mine too (and intrigeri's, I'm sure :)). Sorry for frustrating you like this about something that is obvious for you, but it indeed gets hard for us when we need to consider things we have little or no clue about. And I'm sure you understand we cannot just blindly trust and implement any suggestion we get.
Cheers!
