From the point of view of BCR's blockchain security, if a better and more efficient mechanism can be employed, it should be.

From your point of view, you can acquire as much BCR as you did before while mining something else too if that's what you like. You say that you are not mining at a loss, but if you aren't selling any of what you mine then you certainly are not making any profit either. Mine whatever PoW currency is most profitable, and use some of those profits to buy more BCR? This is a win for you, surely?

It's a pretty wild shift from the crypto 'norm' but I think that PoW norm is fundamentally flawed, and in the long term shifting tx validation to the BNs and adopting a much simpler and easier method for anyone (that other 99.9999999% of plane Earth that will never be able to mess with all the technical voodoo of mining) to acquire BCR and thus access to future platform services, while directly helping to fund development, will be of great benefit to BCR. That future platform utility is what will bring real increased value to your current holdings.

This is a Giant Step for BCRkind in a whole lot of ways. I hope you can come to see that.  

How about BN that mines the block gets the same reward as the rest?

Giving you the chance to acquire BCR more cheaply and easily is limiting you how? If you like you can send me the extra money you would have spent on mining it.

Eh? The block reward may decrease per block but there will be more frequent blocks, plus the current 5 + 5 BCR each block will go, so bidders will actually get more compared to what miners get now.

And since bidding is likely to work out cheaper than mining (for those who don't get free electricity) anyone who wants to can acquire more, more cheaply, more easily. What's being discussed will improve distribution accessibility, not limit it.

It's PoW that's limiting and exclusionary to the 99.9999999999% of the world's population that can't afford the hardware and the running costs of PoW, and don't have and never will have the know how anyway.

Reducing the collateral requirement would only superficially (at first) increase security, because let's face it we'd all just be running 3/4/5 BNs on the same machine(s) instead of 1. We'd have ~100 BN's immediately with a 50k collateral each but no real world improvement. It would however make it easier for new players to get on board going forward.

eg. 50k BCR @ 1000 sats is 0.5BTC - hardly a huge investment, so might attract more people. Those of us who are prepared to invest a bit more will in time be better off as we'll also have BN capital for p2p loans, backed asset issuance etc. It isn't a deal breaker for me either way, maybe it's worth lowering to 50k just to try and ignite more interest. There is still going to be inflationary pressure from new mintage, making BNs more affordable is a good counter to that, but blockchain security should be the paramount consideration.

Just 200 BN's within a month or two, reasonably distributed, would probably give us much better actual security than any pool-mined PoW currency even without per-block subsetting. Subsetting 10% of them would be even better, 160 compromised nodes out of 200 would only give an attacker a 1.1% chance with 20 random nodes needed for consensus. And it just gets better from there.


The block time thing isn't a big issue for me. Faster is better, 30 secs would be fine if lower than that would cause issues. Maybe the new scheme would avoid or mitigate the problem dragos mentioned with p2pool, I'm happy to let those better qualified decide what's technically appropriate.


Very much agreed here on sensibly spending the 6 million, it's no use sat there doing nothing. There's a lot of competition in this space and time is of the essence. And at some point we should set up a BCR testnet.

Speaking of adoption, that will happen as BCR grows into a viable and capable platform, but I frankly don't care very much about 'adoption' by bitcointalk users.    Business to business automated escrow, reserve-backed assets, p2p lending once the credit scoring system is worked out, etc. ... things like this matter more to me.

I know a lot of people don't take the risks associated with pooled PoW mining seriously, or even concede their existence, but there's a reason big money is shy of BTC - ask yourself if you had $millions++, would you trust it to a system that could be taken down and wrecked by compromising just 3 publicly known or easily identifiable machines? That's not security, and just because nobody has bothered to break BTC yet doesn't mean it will never happen, because for any serious adversary it's a cheap trivial op. Having an actually secure blockchain is important.

What's been talked about here recently can
a) Give us actual blockchain security, not just security theatre
b) Fund faster and better development

It's been a very good few pages. 

Taking BTC and other currencies in exchange for newly minted BCR also provides liquidity/reserves in those other currencies for the BNs to provide the holy crypto grail: a decentralised exchange...  

This is brilliant.   

Would bidding be blind, or would the wallet report the current bids to all? The latter would be better IMO.

It doesn't need to be ebay-complex, just a simple Time Left counter, a Current Bids list, a Bid Amount field (estimated from current price) and an Enter Bid button. Oh and a Set Repeat Bid checkbox, regular 'miners' don't want to be doing this manually all the time. Name your price and it'll keep buying for you until you're out of ammo.

Before this can happen though BCR would need cross chain/coloured coin/other currency/fiat? capability - bidding in BCR for BCR doesn't make sense.

And what happens if nobody bids?
No mintage - prolongs emission schedule and should in theory reduce sell pressure, quasi-deflationary...
Mintage burned - keeps emission schedule, definitely deflationary...

I'm still not sure about the ramifications of all this, but if it can be made more secure than PoW (shouldn't be too hard) and gets rid of pissing money away on electricity (the leeches who get 'free' electricity will be annoyed, but somebody is paying for that wasted wattage) then it's an improvement. Aren't KnC building their own hydroelectric dam to power their acres of ASICs? Pretty sure that wasn't a part of Satoshi's original vision of a decentralised currency.

Plus, vastly better tx throughput.

edit: I think someone mentioned it before, but basically this would mean that the money miners spend currently on running costs would be going to... the great BCR Banknode Reserve... bitcreditscc's (DA)C... some communally owned asset held on the blockchain?

"Miners - support BCR, not AMD/Nvidia and the power company!"

Currently small numbers, sure, but if 'mining' BCR involved nothing more than having a wallet online (at a unique IP, or some other method of stopping people running a thousand wallets at a time, like a min 1000 BCR balance) and sparing the occasional CPU cycle to provide service, a lot more people might do it?

This is pretty radical and I'd need some time to think about it, but I like the idea of 1339 transaction blocks per day and 1 minting block. Remember there doesn't need to be any 'hashing' at all by the BNs, they just need to arrive at signed consensus, so waste would be almost zero - how much CPU power does it actually take to process a few (or a few thousand) numbers in a distributed db or RAMpool? - Very little, by modern standards. The lowest end smartphone is overpowered for the job.

However with such a relatively small number of BNs it wouldn't offer much security? Mind you, at any given time, for most PoW currencies, control of a maximum of 3 pools (just 3 machines, run by 3 people, running on publicly known IPs somewhere in the world) are needed to gain >50% of the total hash and create havoc and mayhem... that's the (appallingly low) security target we have to beat.

I have suggested in the past with DRK that only MNs 'mine' - this would work with 2500+ MNs, as a different random subset of them could be used each block for consensus and provide great blockchain security, but with so few Banknodes, not so much. Since BNs and regular wallets are exactly the same from a software perspective, why exclude regular users from the consensus process? This would provide a greater pool of nodes to subset from and would be completely transapernt to users, they aren't going to even notice a few extra CPU cycles now and again.

The idea that the (fixed number of?) minted coins are distributed each day in proportion to whoever has 'bid' for them is intriguing... it would make it a kind of tender/lottery system. How would this interact with the existing market/exchange price discovery system? Might change the price pressure - either up (demand) or down (miner instadumping) - in interesting ways... I still think it would be easier to just dish out the mintage to the selected consensus-subset each block.

I had a similar idea for how BNs could provide perfectly anonymous transactions without any need for slow, complex DS mixing, but I need to work out some details...

PoW is never going to scale to global levels. A globally used currency is by definition going to have a very large market cap and needs to be completely robust against blockchain corruption/manipulation/vandalism/terrorism/etc. PoW is exclusive of poorer users, to whom the concept of spending money on otherwise useless to them computer hardware and furthermore having to pay to run it is justifiably alien.

PoW is mindbogglingly inefficient. Pretend for a moment you know nothing about crypto. Then try selling this cunning plan to yourself:

Current BTC hashrate = 368 PH/s = 368 * 10^15 H/s ~ 4.784 * 10^20 32 bit operations/sec. *  Consuming dog knows how many megawatts of power. To obtain network consensus on a mighty theoretical maximum of 7 transactions per second. Face it, it's toytown crap.


Crouton's PoService Proposal:

Each block BNs (or the network as a whole) choose a subset of miners. Only those assigned miners out of all connected miners validate transactions for that block. All mining logic is the same, just without the wasteful algo make-work, carried out by that subset. Plus maybe a control group or three for extra security. Consensus among subset reached, block gets baked into the chain. All conflicting txes get rejected.

Furthermore, all connected wallets with a unique IP address (and maybe some small minimum balance) are miners. (To avoid people firing up ten thousand wallets each.) Doesn't need to be a fixed IP as miners can be identified on the network with a unique privkey, same as BNs are. And no expensive hardware or the ability to pay running costs are needed because there's no hardware arms race any more. All connected wallets are chosen amongst each block, The chosen subset does what miners usually do, but without the need for megawatts, bake that block and distribute to all. Next block, new subset, etc.

We would have
1. Near perfect decentralisation of mining.
2. Impregnable Blockchain security. Forget 51% attacks. With a subset of eg. 20 miners required for consensus, any attacker would need to control over 80% of running wallets to have a 1.1% chance of a double-spend etc.**
3. Inherently higher tx capacity as it should be faster to achieve consensus among a smaller set of nodes.
4. No wasteful make-work.
5. No hashrate fluctuations, stuck blocks etc. as there wouldn't be any hashrate, so currency emission/blocktimes would be perfectly predictable.
6. Steady and predictable miner income, the same way BN payments should work on a FIFO basis, while avoiding the p2pool dust problem as only elected miners are paid each block.
7. Inclusion - everyone gets a share, not just hardware junkies with cheap/free electricity and BN ops.
8. Peace on Earth.

Now, tell me why this wouldn't be awesome.***



*source:
Quote from: gmaxwell on November 19, 2011, 06:26:11 PM
Ngzhang says above that there are about 1300 32-bit operations (in one bitcoin hash) which sounds about right. So one 1GH/s would be equal to 1,300,000,000,000 operations per second.

**Chance of successful attack ~ (Number of compromised nodes / Total number of nodes)^(Number of nodes in the consensus validation chain)
eg, 80% compromised nodes on a 1000 node network, with a random 20 nodes in agreement needed for consensus gives (800/1000)^20 = 0.011529 ~ 1.11% chance of a successful attack. Whick beats the pants off 51% of compromised PoW miners giving a 100% chance of wreaking blockchain mayhem.

***The usual argument against this is, 'mining gives value to the currency.' To which I reply:
a) Unless your electricity is free, most mining is unprofitable and most currencies trade below their mining costs most of the time, so no, mining doesn't really give value to anything.
b) More importantly, nobody in the real world gives a damn how much it cost you to produce your digital tokens. What they care about, and what ultimately lends value to a coin, is it's utility. When you go travelling, do you care how much it cost The Land of Far Away to print their banknotes? Do you even care much what the exchange rate is? No, you just buy as much as you need and go spend it on bar girls and tequila wholesome and tax deductible things and postcards home.



And PoService != PoStake. You are not staking anything, you are not earning interest based on your balance. You are providing a service - consensus validation of transactions - and getting paid a little income for that service.

I'm not suggesting this be implemented overnight, but in the longer term PoW just isn't going to cut it. Near(er) term, if any algo change or whatever was being considered, I really like SPR's pool resistant solo-only approach. It's not perfect but it's by far the best PoW solution currently available IMO. And it could be slotted right in in fairly short order.

I hear what you're saying, but it would involve a lot of wrangling... How much should each type of node 'cost?' How should the block reward be divided? How much in the way of resources/running costs would each type incur? How long would it take extra to get all these discrete types working in harmony - there are always unexpected problems.

I think it would be difficult to come to a good decision on these things until we see them all in action. Maybe it would be better to get everything running on 'standard' BNs first to evaluate things and then consider subdivision? Might make a great price/fundraising bump in the future too, although that should remain a secondary concern.

I don't know, just voicing my thoughts. But I think the, 'BNs are cheap now but might be more expensive in the future' isn't sufficient reason by itself - nothing is stopping anyone from grabbing enough BCR now. And with working blockchain assets, standard BNs (or the BN network as a whole) can for example 'host' anyone's DAC for a hopefully reasonable fee.

Just a functional DHT running on the BNs with in-wallet access offers so many possibilities... add working IBTP/coloured coins and BCR can do everything Coinbase does and far more... someone remind me how much the Coinbase IPO/ICO thingy raised...?

Under 10 BTC for a share of that starts to look like one hell of a bargain.


edit: someone kick the blockchain, it's stuck again. Don't let me get started on how crap, exclusionary and unscaleable PoW is and how pure PoService would be literally 100000+ times more secure and almost perfectly decentralised.

edit2: blockchain moving again!

Some quick thoughts:

30 secs sounds good to me, no objections.

We could royally piss off the Spreadcoin guys and call them Servicenodes.    

bitcreditscc has mentioned a company formation... maybe this needs some more discussion - you are right, if we all had a better idea of where we were going exactly then dissemination of information would be easier.

I'm with bitcreditscc on this. Particularly at current prices, I don't think a BN is prohibitively expensive at all. Costs about the same as a 5x750Ti mining rig... maybe less if you're patient with your buys?

I've said before, I'm happy for him to sensibly spend the 6 million if it helps. On this subject though I might as well mention the Automated Cronyism By Blockchain that DASH is implementing - done right, with actual community power to approve and veto spending proposals, something like this could be a useful tool for funding future development. But the DASH implementation is shaping up to be pure gravy train for the insiders, and MN owners have no way to block or meaningfully vote against garbage spending as their system is currently proposed.

LOL, "hard coded" - either whoever wrote their platform isn't answering their calls and they don't know how to do anything themselves, or cex.io is actually written in Concrete++ and running on Intel's latest Tectonic Plate architecture.

Neither of which bodes well if they encounter an actual problem. 

Have you looked at it though...  private messaging, assets, distributed storage via DHT for a start - that's going to make BCR one hell of a versatile platform.

Yeah, might take just a little while to fully integrate and debug... 

Yeah it's a pointless and initially irritating change, and you'll need to update your install/update scripts.

Innovation, BTC style!

BN block reward up on schedule without issues.

Pull request #29 awaits you!

I've gone with what to me seemed a clean(er, anyway) and consistent look, but it can all be changed and I'm open to suggestions.  UI design isn't my day job.  

Got the Banknode Manager page working again:



Haven't tried creating one via it yet though.

Fiery rattlesnakes.

https://github.com/thelonecrouton/bicreditsnew/commit/38eb56d3243010bafd1634498383a31f9022c46b
https://github.com/thelonecrouton/bicreditsnew/commit/f0a2dd1116249419ac7ef28843437fa42e47b3fe