[bitcoin-dev] summarising security assumptions (re cost metrics)

[Eric Voskuil]

On 11/06/2015 03:41 PM, Chris Priest wrote:
>> The bigger point however, which Erik explained, was: widespread use of
>> APIs as a sole means of interfacing with the blockchain also
>> contributes to reducing network security for everyone, because it
>> erodes the consensus rule validation security described under
>> "Validators" in the OP.
> 
> I completely disagree with this. You are implying that there is some
> sort of ideal ratio of full nodes to 'client only' nodes that the
> network must maintain. You seem to be implying that if that ideal
> ratio is to somehow be disrupted, then security suffers. My question
> to you is what is that ideal ratio and what methodology did you use to
> come up with it?

Nobody has advocated a golden ratio.

> The way I see it, the security of the system is independent on ratio
> between full nodes and lightweight nodes.
> 
> In other words, if there are 100,000 lightweight wallets to 100 full
> nodes, then you have the same security profile as one with 100,000
> full nodes to 100 lightweight wallets.

This is a false dichotomy. Both scenarios are poor for security, as
nobody with a wallet is validating. It's entirely possible, even
probable, that one person controls all of the nodes.

> I think most 'big blockers' think the same way I do, hence the rub
> between the two camps.
> 
> Small block people need to make a better case as to how exactly full
> node ratio relates to network security (especially the 'for everyone'
> part), because the link is not clear to me at all. Small block people
> seem to take this simple fact as self evident, but I just don't see
> it.

Fewer people independently validating their own transactions means trust
is placed in fewer people. The degenerate case of one validator and
everyone trusting it is dispositive, and equates roughly to the Federal
Reserve.

> On 11/6/15, Adam Back <adam at cypherspace.org> wrote:
>> You're right that it is better that there be more APIs than fewer,
>> that is less of a single point of failure.  It also depends what you
>> mean by APIs: using an API to have a second cross-check of information
>> is quite different to building a wallet or business that only
>> interfaces with the blockchain via a 3rd party API.  There are
>> different APIs also: some are additive eg they add a second signature
>> via multisig, but even those models while better can still be a mixed
>> story for security, if the user is not also checking their own
>> full-node or checking SPV to make the first signature.
>>
>> Power users and businesses using APIs instead of running a full-node,
>> or instead of doing SPV checks, should be clear about the API and what
>> security they are delegating to a third party, and whether they have a
>> reason to trust the governance and security competence of the third
>> party: in the simplest case it can reduce their and their users
>> security below SPV.
>>
>> The bigger point however, which Erik explained, was: widespread use of
>> APIs as a sole means of interfacing with the blockchain also
>> contributes to reducing network security for everyone, because it
>> erodes the consensus rule validation security described under
>> "Validators" in the OP.
>>
>> Adam
>>
>>
>> On 6 November 2015 at 09:05, Chris Priest <cp368202 at ohiou.edu> wrote:
>>> On 11/5/15, Eric Voskuil via bitcoin-dev
>>> <bitcoin-dev at lists.linuxfoundation.org> wrote:
>>>> On 11/05/2015 03:03 PM, Adam Back via bitcoin-dev wrote:
>>>>> ...
>>>>> Validators: Economically dependent full nodes are an important part of
>>>>> Bitcoin's security model because they assure Bitcoin security by
>>>>> enforcing consensus rules.  While full nodes do not have orphan
>>>>> risk, we also dont want maliciously crafted blocks with pathological
>>>>> validation cost to erode security by knocking reasonable spec full
>>>>> nodes off the network on CPU (or bandwidth grounds).
>>>>> ...
>>>>> Validators vs Miner decentralisation balance:
>>>>>
>>>>> There is a tradeoff where we can tolerate weak miner decentralisation
>>>>> if we can rely on good validator decentralisation or vice versa.  But
>>>>> both being weak is risky.  Currently given mining centralisation
>>>>> itself is weak, that makes validator decentralisation a critical
>>>>> remaining defence - ie security depends more on validator
>>>>> decentralisation than it would if mining decentralisation was in a
>>>>> better shape.
>>>>
>>>> This side of the security model seems underappreciated, if not poorly
>>>> understood. Weakening is not just occurring because of the proliferation
>>>> of non-validating wallet software and centralized (web) wallets, but
>>>> also centralized Bitcoin APIs.
>>>>
>>>> Over time developers tend to settle on a couple of API providers for a
>>>> given problem. Bing and Google for search and mapping, for example. All
>>>> applications and users of them, depending on an API service, reduce to a
>>>> single validator. Imagine most Bitcoin applications built on the
>>>> equivalent of Bing and Google.
>>>>
>>>> e
>>>>
>>>>
>>>
>>> I disagree. I think blockchain APIs are a good thing for
>>> decentralization. There aren't just 3 or 4 blockexplorer APIs out
>>> there, there are dozens. Each API returns essentially the same data,
>>> so they are all interchangeable. Take a look at this python package:
>>> https://github.com/priestc/moneywagon
>>

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 473 bytes
Desc: OpenPGP digital signature
URL: <http://lists.linuxfoundation.org/pipermail/bitcoin-dev/attachments/20151106/1cfc4dae/attachment.sig>