Gold collapsing. Bitcoin UP.

[VeritasSapere]

@SysMan Unfortunately because of the refusal of Core to increase the blocksize, we need to change the "development style" or what I would call the governance of Bitcoin now. Since the actions of Core are leading us towards a fundamental change in the economics of Bitcoin. I think that their conception and vision for the future of Bitcoin is fundamentally flawed, it represents a divergence from the original vision of Satoshi. All that Core needs to do is stall in order to bring about this change, this is why action is required now. The market will not react well to a Bitcoin that is more expensive and less reliable, Bitcoin must still be able to compete.

 

[theZerg]

@SysMan, I think that aquent's comments about tradeoffs are the most important. If the tradeoff is between Bitcoin users growing 50% this year and full nodes decreasing by 50%, I'd pick user growth any day. Its much harder or impossible to win market share back from an alt-coin once a company has started going down that road, than it is to turn a full node back on.

And this adoption vs full nodes IS your worst case argument. But it is unproven. What if adoption causes price increases which creates more interest and therefore more full nodes are run?


Finally, I haven't read your paper carefully, but I have a hard time believing your claim that 40% of nodes will "drop off" if we go to 2MB. The average web page today is 2.25 MB.

Bitcoin nodes data use is very variable and very much related to the number of connected nodes. So nodes could simply make half of the connections they do today (from 100 to 50 nodes, say) and handle 2MB blocks with very similar bandwidth as today's 1MB nodes. And the network will still be well connected, there's a lot of room there before we get a sparse network.

With Bitcoin Unlimited you can turn on "traffic shaping" which lets you specify max and average upload and download speeds. I was running these at modem speeds 56KB/s no problem, but my client had fewer P2P connections.

 

[cypherdoc]

@SysMan

if Brain @Coinbase really wanted to take over, he'd be advocating for a blocksize decrease to force more tx's offchain into his centralized service. but no, he is actually looking out for what's best for Bitcoin; growth worldwide. which, btw, will actually make his biz explode even larger.
[doublepost=1457123073][/doublepost]from Brian:

In my opinion, perhaps the biggest systemic risk in bitcoin right now is, ironically, one of the things which has helped it the most in the past: the bitcoin core developers.

 

[AdrianX]

Ok may be it when time to calculate block time delays on release ? just do it, fill free...
SPV is pure benefit there are nothing about network safety, avoid orphans or similar... just monetary profit.

To you there is just benefit and it's not about network safety. That's because how the bitcoin incentive system is designed. But in fact Miners are incentivised to follow their interests ("[mining] SPV is pure benefit"), the result is greater network safety with bigger blocks. see this paper:
http://www.bitcoinunlimited.info/resources/1txn.pdf
[doublepost=1457123666][/doublepost]

@SysMan

I was very happy to see that you've joined this forum!

The only thing I want to add to @cypherdoc's points here is to stress that we wouldn't be "firing Core" and "all moving to Classic." Classic is just a vehicle to voice our preference for larger blocks. Once the Classic hard fork looks inevitable, either:

(1) Core will make a change so that Core nodes track consensus,

OR

(2) The majority of developers currently working under the "Core brand" will join Classic, Unlimited, or form new projects.

If you don't want to run Classic, then why not just recompile your Core nodes so that they produce version 0x30000000 blocks (and thus count towards 2MB activation)? If the hard fork activates, then during the 28-day grace period you could upgrade to remain compatible with consensus.

I'd like to add to Peter's point, once the Classic 2MB fork is activated Core and its developers will not just quiet and leave bitcoin. What is most likely is those employed by Blockstream will carry on with their roadmap and accommodate the hardfork to 2MB.
[doublepost=1457124411,1457123329][/doublepost]

That's the point.

If Classic is a coup or a takeover, then it proves that it's needed because one group of people believe they are the legitimate rulers!

If people stop buying iPhones and start buying Android phones instead, nobody calls this a coup, because nobody believes that Apple has the right to demand that people only buy their phones.

If Core believes that other people to choose other software is coup, then they believe themselves to be the rulers of Bitcoin. For holding that belief, they are not qualified to be Bitcoin developers and every responsible business should stop using their software.

This doesn't mean use Classic necessarily - it would be even better for more people to start running their own implementations.

I just wanting to reiterate this point. thanks @Justus Ranvier

 

[Roger_Murdock]

That's the point.

If Classic is a coup or a takeover, then it proves that it's needed because one group of people believe they are the legitimate rulers!

If people stop buying iPhones and start buying Android phones instead, nobody calls this a coup, because nobody believes that Apple has the right to demand that people only buy their phones.

If Core believes that other people to choose other software is coup, then they believe themselves to be the rulers of Bitcoin. For holding that belief, they are not qualified to be Bitcoin developers and every responsible business should stop using their software.

This doesn't mean use Classic necessarily - it would be even better for more people to start running their own implementations.

Exactly. I think this gets at one of the biggest "psychological barriers" to raising the block size limit. Even many of the people who understand that the market is ultimately in control of Bitcoin's direction appear to conceptualize Bitcoin as a kind of "representative marketocracy." So, from that perspective, forking to Classic means voting the "Core Party" out of power and electing the "Classic Party." But that sounds like a very significant and potentially scary change. "Is Classic really ready to lead?" But a healthier view of Bitcoin's governance would see Bitcoin as something closer to a "direct marketocracy." Again, every line of code put out by any development team is a separate offering that the market can accept, reject entirely, or modify. By definition, Core's suggested 1 MB block size limit is just that, a suggestion. Declining to follow that particular suggestion via a simple code change like that provided by Classic is not, or at least should not be, some hugely momentous "coup." If I had to use a political metaphor for what Classic is attempting, it's a lot less like a coup and a lot more like a line-item veto.

 

[SysMan]

@SysMan, I think that aquent's comments about tradeoffs are the most important. If the tradeoff is between Bitcoin users growing 50% this year and full nodes decreasing by 50%, I'd pick user growth any day. Its much harder or impossible to win market share back from an alt-coin once a company has started going down that road, than it is to turn a full node back on.

And this adoption vs full nodes IS your worst case argument. But it is unproven. What if adoption causes price increases which creates more interest and therefore more full nodes are run?


Finally, I haven't read your paper carefully, but I have a hard time believing your claim that 40% of nodes will "drop off" if we go to 2MB. The average web page today is 2.25 MB.

Bitcoin nodes data use is very variable and very much related to the number of connected nodes. So nodes could simply make half of the connections they do today (from 100 to 50 nodes, say) and handle 2MB blocks with very similar bandwidth as today's 1MB nodes. And the network will still be well connected, there's a lot of room there before we get a sparse network.

With Bitcoin Unlimited you can turn on "traffic shaping" which lets you specify max and average upload and download speeds. I was running these at modem speeds 56KB/s no problem, but my client had fewer P2P connections.

- 1st you can not be sure about 50% increase of full nodes.
- 2nd Please then read.
2Mb blocks is not what I'm afraid of... but even they will push out some nodes, and 4Mb will push even harder and more because of p2p/traffic/cpu/mem/disk usage. there are multiple parameters and numbers say be them-self precise.
- 3rd opposite of your said - less full nodes left , more connection they should support to support also SPV nodes. meaning number of their connection & load will grow in percentage of nodes decrease.

p.s. forget about 56Kb, 2Mb block will require for full node 8...10+Mbit line at least + . for miners even bigger, this push out most of DSL users.

 

[cypherdoc]

@SysMan

First off, your results differ greatly from those done by Gavin that showed 20MB blocks would be OK.

Second, no where in what you've said is there any estimation of the economic effects of what will happen with 2MB blocks. I think by doubling the capacity throughput of txs the price could easily double. Along with that, it would make the cost of running my full nodes net neutral. Miners and merchants would be making more and they also would be encouraged to run their own full nodes.

This growth dynamic from 0 to 1mb is precisely what we've seen since 2009 so why change things with radical SW?

How does BF plan to make money as rewards decrease long term considering you're fine with SW pushing txs offchain to LN?

 

[Peter R]

@theZerg, @SysMan, @cypherdoc

A similar study to BitFury's is this recent paper:

By measuring block propagation to nodes, the authors conclude that 90% of the network could keep up with with an average block size of 4.1 MB, while it would take 37 MB blocks (on average) to knock out half of the network. So the results are similar to @SysMan's, but a bit more ambitious too.

What is really great is that Xthin for Bitcoin Unlimited is designed to target exactly this bottleneck.

EDIT: @SysMan: just wanted to say that BitFury has been doing a great job at sharing its research with the community via the various papers that it writes! Please keep it up!!

 

[Matthew Light]

Undeserving? Do you have any idea how hard it is to HODL through all the ups and downs in Bitcoin's history? See here to understand about power law distributions and the psychological difficulty of holding through million-percent growth, and here for why Bitcoin's mature ledger state is so important.

Cryptocurrency is not a religion, it's a technology.

When I say that many (not all!) of the current large HODLers are undeserving - well look at their behavior. Censorship, DDoS, choosing Bitcoin Core over Classic, as well as a quite collection of grifters and scammers who have basically stolen their Bitcoin wealth.

Of course there are many who have acted admirably.

But nobody owes any Bitcoin HODLers anything.

All I can say to you is keep your eyes open and pay attention, and don't miss out if another ledger looks like it is going to pass Bitcoin's.

 

[freetrader]

Finally, an(other) honest way to make money

London Blockchain Security and Financial Cryptography Research Competition and Research Prize Fund 2016

http://www.nicolascourtois.com/bitcoin/blockchain_UCL_London_challenge_2016.pdf

"Participants from business and start-ups are allowed, however blockchain professionals and well-funded professional researchers will not qualify for cash research prizes unless we decided otherwise."

Glad they left themselves an out!
Honi soit qui mal y pense ;-)

 

[cypherdoc]

@theZerg, @SysMan, @cypherdoc

A similar study to BitFury's is this recent paper:

By measuring block propagation to nodes, the authors conclude that 90% of the network could keep up with with an average block size of 4.1 MB, while it would take 37 MB blocks (on average) to knock out half of the network. So the results are similar to @SysMan's, but a bit more ambitious too.

What is really great is that Xthin for Bitcoin Unlimited is designed to target exactly this bottleneck.

EDIT: @SysMan: just wanted to say that BitFury has been doing a great job at sharing its research with the community via the various papers that it writes! Please keep it up!!

i'd also love to hear our own @davecgh comment on btcd's research where they found that 100MB blks could be tolerated by the current network.

 

[79b79aa8]

I was present at this event yesterday: http://digitalchamber.org/dcsummit.html. In attendance, among many others: Jeff Garzik, Vitalik Buterin, Sam Cole (KnC Miner), Stephen Pair (BitPay), Blythe Masters. People from FBI, FCC, Homeland Security; Citibank, Nasdaq, Deloitte, etc. Also Jamie Smith from BitFury (@SysMan -- excellent to have you here).

Here is a short report of information that surfaced that seems most salient to me (but if interested feel free to ask questions and I will try to answer).

1. Private blockchains (permissioned ledgers if you wish) are here now. IBM is launching a massive initiative, to be marketed to their industry clients, with the purpose of helping them streamline business end to end. Database security not discussed. Masters' company (Digital Asset Holdings, focused on wholesale financial services), is building a ledger for the Australian stock exchange. Nasdaq developing their own. R3C is well on its way. Startups like Chain, Bloq, Blockstream are consulting and developing software for this space.

2. Perceived benefits of the above vs. using the bitcoin blockchain: (a) ability to directly digitize existing asset classes and put them on a shared ledger, without going through a network-issued currency; (b) privacy: some transactions not suitable to be encoded in a public chain, e.g. ones that would reveal sensitive info. to your competitors, similarly for contracts; (c) sense of risk/taboo involved in adopting private currency (d) legal and practical obstacles to settling trades of securities and similar assets on a public chain whose security is provided by unknown parties in many jurisdictions (miners).

3. Sam Cole of KnC said his company was already consulting with businesses on how to set up mining operations to secure these private blockchains. I asked privately for a bit of elaboration. Said he: Step 1: securing servers for an in-house, single institution blockchain (presumably against attacks by their own IT). Step 2: setting up mining rigs for a multi-institution permissioned ledger (like R3C). This involves schemes such as geotagging mining data so that participants can ensure each party's mining is done in-house and does not exceed pre-determined hashing limits, in order to prevent 51% attacks. I said this seems silly. He agreed, but said this is what businesses are prepared to do at this point with this tech. In the long run, he expects them to fall back on the open protocol (gave an analogy with SMTP, and how institutions attempted to create their own systems only to eventually return to the original standard). Right now he'll take the business.

4. Asked Garzik privately about blocksize limit. Said he: (a) limit will increase (b) but bitcoin blockchain, no matter how it scales, will not be able to hold all transactions in the world. So alternative chains will arise, some permissioned (this has already happened as per 1), some interoperable with bitcoin blockchain. (c) It is necessary to establish a bitcoin kernel (d) many forces at play right now trying to influence how this gets settled, Blockstream is one but equally powerful others are counterbalancing. (d) His conclusion: market will work things out.

PS. @cypherdoc, I feel it is getting hard to read you with so many acronyms and abbreviations. Twitter is crimping your style.

 

[Matthew Light]

Undeserving? Do you have any idea how hard it is to HODL through all the ups and downs in Bitcoin's history? See here to understand about power law distributions and the psychological difficulty of holding through million-percent growth, and here for why Bitcoin's mature ledger state is so important.

Regarding your second link, a key quote from it:

"Simply put, a good investor can predict the future better than other people can."

Fully agree.

My prediction right now is that Bitcoin is heading for irrelevancy and that Ethereum will take over as the primary cryptocurrency platform. If i am right about that prediction, I will have greatly amplified my financial wherewithal. If circumstances change and Bitcoin picks itself up off the mat, I hope I will be wise enough to change my investments quickly enough to do well.

 

[_mr_e]

I'm sure you will, and when bitcoin proves that the honey badger truly don't care, there will like be an massive ether dump and an epic btc bull run!

 

[majamalu]

Hey @cypherdoc , let me know you what you think about this idea (please note that I'm open to the idea that this is completely silly):

The chinese miners want the btc price to go up, and they seem to believe that the Core "roadmap" is the way to make it happen. What if we pool some btcs in order to prove them wrong with a simple bet?

We can say -- perhaps to the AntPool guy: If you start mining classic blocks right now and keep doing it for a couple of weeks, I bet btc the price will go up at least USD 100 during that time frame. If I loose, I will give you 10 bitcoins; If I win, you will only have to give me 5 bitcoins. Or something along those lines.

That could be a good way to make a point, even if they don't take the bet.

 

[AdrianX]

PS. @cypherdoc, I feel it is getting hard to read you with so many acronyms and abbreviations. Twitter is crimping your style.

PPS (Post Post Scriptum) @cypherdoc and everyone for that matter the info discussed here is very interesting and a lot gets lost in the acronyms and abbreviations.

I think it's worth taking the time to define the acronyms as (words) every few pages. Ultimately you're not preaching to the converted but the interested and these ideas are too valuable to not understand.

It's the huge following of the old Gold Collapsing thread had that triggered its closing. I'm still confident the readership will pick up and I would like to see ideas expressed in any way that makes them easier to understand for everyone (especially for new readers.) and me.

 

[Roger_Murdock]

Something I've been mulling over lately. Let's say that you were a very early Bitcoin adopter with a huge stash and you were worried that Core's mismanagement is beginning to pose a serious threat to your investment. How could you best use your available war chest to try to steer Bitcoin back on track and how much would it realistically take to have a meaningful impact?

Obviously, you could try buying up a big chunk of hashpower and mining with Classic. But I don't really have any sense for how much it would cost right now to acquire say, a quarter of network hash power. (And would that be enough to trigger a tipping point?)

Another idea would be to try your hand at bribing one or more key pool operators / Core devs. But how many and which ones would you need to bribe? And what's the going rate for something like that?

Another idea would be to throw some money at funding alternative development teams to allow them to hire additional devs / devote themselves full-time to those projects. This might also help these teams to be viewed as "more credible" to the miners who, per my previous comment, imagine that a fork represents some huge change in leadership.

Another thought I had was creating a bounty that would be split among the 750 triggering Classic blocks following a successful hard fork. So if your bounty is 1500 BTC, each of those 750 blocks would get an additional 2 BTC sent to their block reward address. Especially after the upcoming halving, that would be a fairly significant bump, no? But still relatively cheap in the scheme of things (about $620,000 at current prices). As an aside, you might want to set some kind of requirement to make sure that the hard fork has "taken" to reduce the risk of small-block miners attempting to game it (e.g., only award the bounty following the network's creation and acceptance of 1,000 >1-MB blocks).

The last idea I think is pretty interesting. And of course, there's no reason it couldn't be crowd-funded as opposed to sponsored by a single mega-whale. Has anyone else proposed something like this? Could it be "smart-contractified"? Do you guys think it's a plausible idea / have thoughts on how to implement it / how big the bounty would need to be?

 

[SysMan]

I was present at this event yesterday: http://digitalchamber.org/dcsummit.html. In attendance, among many others: Jeff Garzik, Vitalik Buterin, Sam Cole (KnC Miner), Stephen Pair (BitPay), Blythe Masters. People from FBI, FCC, Homeland Security; Citibank, Nasdaq, Deloitte, etc. Also Jamie Smith from BitFury (@SysMan
-- excellent to have you here).

Next time ... this time was busy. someone should work also time from time

 

[AdrianX]

Regarding your second link, a key quote from it:

"Simply put, a good investor can predict the future better than other people can."

Fully agree.

My prediction right now is that Bitcoin is heading for irrelevancy and that Ethereum will take over as the primary cryptocurrency platform. If i am right about that prediction, I will have greatly amplified my financial wherewithal. If circumstances change and Bitcoin picks itself up off the mat, I hope I will be wise enough to change my investments quickly enough to do well.

Nice, I predict we don't actually know but the likelihood of Ethereum overtaking bitcoin is still rather slim. I have to agree with @cypherdoc "most people in this space are going to lose money™" I think the ETH could still climb a lot as bitcoin gets #rekd by the Core developers, but I still think bitcoin is going to be the one to own at the end of the day.

 

[SysMan]

@theZerg, @SysMan, @cypherdoc

A similar study to BitFury's is this recent paper:

By measuring block propagation to nodes, the authors conclude that 90% of the network could keep up with with an average block size of 4.1 MB, while it would take 37 MB blocks (on average) to knock out half of the network. So the results are similar to @SysMan's, but a bit more ambitious too.

What is really great is that Xthin for Bitcoin Unlimited is designed to target exactly this bottleneck.

EDIT: @SysMan: just wanted to say that BitFury has been doing a great job at sharing its research with the community via the various papers that it writes! Please keep it up!!

https://panteracapital.com/wp-content/uploads/4a3d58986cc386bdf8e95ff81d4b0760-the-bitcoin-p2p-network-.pdf

Once again not the only _block_ propagation is bottleneck.
transactions propagation, organization of p2p connections, connection handling, connection bandwidth controls, dnsseed/prober are all bottleneck's.

Right now network from old nodes bitcoind 0.7.x/0.8.x and old style "high S","dust","low.fee" are flooding the network from new version nodes 0.11+ decreasing p2p network efficiency for at least 3-5% (and I may explain exactly how and why).