Coming in a bit later here. To me the essence of the reasoning for the fork is to head off the possibility that sometime in the not too distant future the demand of bitcoin users to transact on the blockchain will exceed the supply. That certainly might happen if fees don't respond to supply-demand economics. Those pushing bitcoin to compete with MasterCard/VISA are, IMHO, a bit crazy. Like shoemakers always reaching for a hammer every time these people see a transactional opportunity they reach for the blockchain. I think that's just plain silly. Bitcoin is not well suited for all transactional situations. In the longer-term it seems great to replace, for example, SWIFT and bank wires but rather poor for those that require cheap- or ultra-cheap real-time settlement whereas some alternatives seem tailor-made for this. Then again maybe I am missing the key reasoning for this fork. On Fri, Jun 19, 2015 at 5:07 AM, Dr Adam Back <adam@cypherspace.org> wrote:
Its clear Gavin knows more about Bitcoin code and detailed micro algorithms than I do (there are many detailed algorithms for anti-DoS etc at code level which I do not know).
Its possible I know more than Gavin or have a better internalised reasoning about the logic and design parameters for about decentralised systems and distributed trust systems, and ecash protocols, threat models in p2p privacy systems - which is quite a big slice of what Bitcoin is trying to do. Or not - I dont know all of Gavin's expertise nor career experience! Something you may not realise is a bunch of us on the cypherpunks list back in like 1995-2005 spent a lot of applied research effort into finding a way to do something with the characteristics of bitcion. My PhD is in distributed systems also.
Anyway I do not mean to have claims to authority, particularly because I believe firmly in pure meritocracy philosophically and detest such argumentation as a failure of reason, but coincidentally I do actually know something about it and worked on it on Bitcoin-like system design and p2p novel trust-model & security model on and off for 20 years.
But I do think people who are proposing big-blocks are underestimating and being super-optimistic about a range of things, almost to naive extent. I am not imputing unsaid things, Gavin wrote many blog posts on these topics. Mike Hearn made some videos and posts about his views, and they are quite disconnected from p2p privacy system design thinking. Someone should probably respond to some of those posts to clarify why they think some of these assumptions are incorrect and optimistic to prior experience and precedent.
Adam
On 18 June 2015 at 20:24, Sean Lynch <seanl@literati.org> wrote:
On Wed, Jun 17, 2015 at 3:51 PM Dr Adam Back <adam@cypherspace.org>
wrote:
If people on *cypherpunks* cant get the points in the post, I think the world has a problem. The price of security in a distributed system like bitcoin is eternal vigilance, but if people dont understand what constitutes a risk and hence what to be vigilant for, the meta-system can be unreliable and lose its assurances. I think we need to explain some more concepts and probably people will over time learn things and and an influencer pyramid emerge as happened in privacy technology.
Yes, I'm sure that when people who disagree with you, it's always because they are wrong and never because you don't understand the situation as well as you think you do. I'm sure you know more about Bitcoin than Gavin does.