Topic: ◈◈Bitcredit ◈◈ Migrating to UniQredit◈◈ - page 140. (Read 284545 times)

bitcreditscc am i right in assuming that your choice of active BNs comes off the fact that they are almost always online ? Local wallets of users are unreliable and the numbers swing about widely, BNs are online at low and at high prices there by creating a known set whose properties we can easily manipulate.

well if people aren't going to participate, then these heavy decisions are going to be made by a small group and it ceases to make any sense to discuss them publicly. One thing i would totally reject is an opinion offered after the fact.

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?

all relevant participation is welcome 

people aren't mining because the perception of overwhelmingly profitable mining is not advertised and is highly subject to price fluctuation.

a subset of 20 seems small , but maybe we should poll total online wallets, i would not be surprised if there are less than 50

Well I am mostly a lurker, but I did my part to address #1. Just got a new BN, and I'm enjoying the 18.75 BCR rewards 

Also I've been CPU mining this coin, and it's been pretty rewarding (lots of small rewards). I wonder why more people aren't mining, since it's an established coin with a masternode-type system, and you can get decent rewards with any CPU.

not enough people participating in these discussions.

1) BNs are too few, we'd need to alter this a little. Also we have to have a system that is secure , no creating loopholes. 

2) i am all for altering some specs to speed up the chain, though this will have to be an update in and of its own.

3) Overall changes to PoW would have to come before any of the other changes if we agree

4) Enacting the changes currently being discussed, would result in us needing to find a way to dispose of the pre-existing 6 million, likely we'll distribute it similarly

I agree, it sounds very light resource wise, as for the number of BNs...we already have a solution , hint hint.

VISA handles on average around 2,000 transactions per second, can you see the shape BCR is headed for? 1440 transactions per second in a decentralized system, which can be scaled up even higher!!!

I think that BNs would be better suited for mining, most are run on pretty good hardware and we use them for more robust reasons than DRK.

As for fixed number of coins everyday, i think it's actually good. Think on it, it's not far different from the actual market in that there are hig and low days. But the one significance is that whoever is getting the coins at the end of the day is actual contributing to development of the ecosystem.

I thought the dev would disappear after a month or two, especially at the highs but he's stayed and he's actively working. In the past year, how many projects still have their original dev, or any dev at all?

Most coins have stalled in development and just wait for pumps , this guy keeps going, i'm sure if he got the support he needs he could push this to the front.

btw..anonymity already exists on the blockchain, these extended versions are gimmicks and truthfully have no commercial application. I'm all for privacy and freedom, but most basic users want to an ABC 123 kind of environment, i introduced my eldest uncle to Bitcoin and BCR last week, do you know what he called the most handy item? The Block Explorer, he siad he had read about BTC but never really understood why someone would want to use money that can't easily be accounted for and proven.

Commercial application is where the real money is, the privacy of just and address that can be changed anytime should be enough for anyone who is not doing nefarious things

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...

How about we replace the proof-of-work with a signature verification for a public/private key pair?

Average blocks per day are 700-800. So that would be 40000 BCR divided among all donors a day? Using the AGS syste on a blockchain would be super!!! Though it has it's quirks.... say on a day when there are no donations where do the coins go?

i assume this funding will be for development and price floatation?

what guarantees would we have on control over development spending and base supported price?

Do you intend to do this polling in client or using additional software ?

More details!!!!
 

Seems we are almost always on the same wavelength, i had a slightly different idea but we are in the same direction.

Signature PoW

I propose a system where only active BNs are miners. Basically instead of a hash rate based PoW, we have a signature based PoService. BNs would therefor be the only waste point for the network. Blocks are produced in mostly the same way except that they require a signature from a BN. A block with an invalid sig is rejected.

Some may say this is elitist, but let me remind you that PoW serves two purposes, distribution and security. The security factor will be dealt with and now only distribution remains. If you were willing to burn $12 (excluding hardware & maintenance cost ) to mine ~ 2000 BCR per day, are you not better off buying 7 000 BCR for the same price? I saw an interesting method of coin distribution used by Bitshares for their (AngelShares). What they did was take donations in PTS and BTC all day, then they would tally the total and divide 10000 AGS between all who had donated. Theirs was an off blockchain method that required trusting the developers

I believe this can be done by SigPoW requiring no trust and thus maintaining autonomy.  Let me expand:-

1) We set up donation addresses for whatever currencies we will accept. (this number can be as high as we want)
2) These addresses are polled once a day and a list of addresses and percentages is generated based on price at poll time
3) Whatever BNs are producing these blocks maintain a list of those who get their payout today
4) at the end of the day, the daily amount of coins is distributed directly to their addresses

Why PoService is superior to PoW

• Highly efficient
• Much faster, less hashrate swings
• 99% reduction in leakage to utility companies
• Convenient for users

The payout can be adjusted as we see fit, "off blocks" are low reward blocks with only tx fee. these will fly by very quickly and are expressly to process transactions in as little time as possible, look at crouton's numbers to see what can be done. We already have 20 MB blocks @ 1 minute intervals so our tx rate would be 1440 tx per second!!!!!!

"payout blocks" would contain transactions and the payouts, we can set these blocks to anytime we want, or any saturation point we want. I would suggest one block a day, as this reduces resource wastage even more because polling would be done only twice every 24 hours.

It would take a bit of doing but this system would propel BCR and all our development giant leaps forward.

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.

Don't want to sound like a broken record, but we really need a solo miner for this coin.  I think network stalls hurt the brand.

Also, having all hash going through the 2-3 pools feels bad as a miner when 10 BCR is lost per block (those +5 and +5 payments every block). Second, mining is a good way to attract people to the coin.  A lot of people wont buy a coin but are willing to boot up their GPU to get a few coins. (I theorize that this is works with the fact that people have bizarre behavior patterns dealing with endowment and risk aversion behavior ..i.e Bell 82' or Kahneman and tversky 79').  Later on the same people who mined, might become buyers or at-least become stabilizing factors.

Bottom line, BCR is having network issue due to the current mining paradigm for the coin, there are many possible fixes (get a solo miner, change the algo to a more supported one, go multi algo etc)(each with pros and cons) but likely this core element needs to be addressed in the near term.

massive drop of in net hash caused the hang up, network already adjusting after solving last block.

One big ant is less effective than many small ants.

I want to work with the market as it is right now, there are no guarantees. I know i haven't been as active a you have (tons of respect) but i fell really strongly about  adoption if we adjust the system now. Rather we decide on the base now, they integrate each new type as it pops up?

Also we are relatively unknown, so if things go as you say, that 10 BTC mark won't last long soon it would be 20 BTC or more, which would make it difficult for smaller entities.

either way, we've both made our points, let the rest speak up.

Meanwhile, about that company has the dev spoken anymore on it? I'd like to see how they plan to structure it and what it means for BCR and development. Maybe we can use it as a proof of concept .

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!

That's the issue, at current price, but a few days after the major updates go live that won't be the case, yet that is the time we hope many people will realize and come aboard the project when it counts. 250K is low right now and most can get a BN if they put their mind to it, but if the price crosses the 10K sats level, smaller guys wont be able to form their companies even i they wanted to. Rather than just 17 guys with 250K BCR , why not have 100 or 200 guys with 50K , then the 17 who want full BNs can use something like "proof of reserves" to become a bank? If it was just Banks then 250 would be reasonable and justified, but with smaller businesses that won't fly. i think something lower would be better. Current BNs could just split their funds into multiple ones if they are worried about their share of the payouts.

I really think a floor of 50K is better, then you can just add additional rules for each level of node. Even the code supports different types of nodes

I saw this amazing little tidbit on the git and i think it can be used for that :-


Each special node would have extra rules.

Ordinary company/service -- 50K
Asset Node - 50K + 100K
Bank Node- 50K + 200K

and so forth and so on. Looking at the code, it does not look hard, we just need to agree and convince the others. In this way you can even run multiple types of services on the same node (unless i missed something)

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.

I have no problem letting people know what i am working on, it's just that the best way to do is is diagrams, and trust me...you do not want to see anything drawn/designed by me.

I can get behind the block rate adjustment. The change in collateral is something else though, the reasoning is not without merit but tbh that adds something new to conceptualize-design-implement to my already long to do list.

This will have to be something that current BN owners agree to. However, i will not agree to anything ridiculous. it has to take into account that we want the investment to be significant an also take into account that companies like a Bank or Insurance will require far much more while still being viable.

Well, the wait is gonna be a bit long. I managed to catch the guy on hangouts and have a face to face chat. Apparently that's just phase one there are 3 phases. 

During our discussion i asked him some questions about the alteration in direction, meganetwork, Colored credits and all the stuff he has been working on. He gave an explanation (very brief) then i read the new draft white paper... I made a few suggestions and he instead asked me to come to the community and speak up so here's what i have.

I'd like to propose some changes :

1) Lower Block time to 15 or thirty seconds - this will decrease confirm times while also pushing up block count to expected 1440 a day instead of the current ~700. This benefits users because transactions will get confirmed faster and BN owners in that their investment gets a faster reward.

2) Re-brand Bank Nodes to a name that reflects that these nodes will not be limited to banks only. Since people will be opening literally any type of company/service they want, we should make sure this is reflected.

3) Put a lot more information out in the public and let investors see more of what he is doing behind the scenes i understand that he wants to perfect things, but if there are three phases and he has already pushed the first, an explanation and community involvement can't hurt.

4) Lower BN collateral. (he flat out refused but i'll put it here any way) Since we are no longer limited to Banks, smaller companies will find that cost prohibitive. Banks can use additional security or additional code that requires extra money , but the base of forming an entity should be lowered.

5) Create a BTC/BCR donation scheme for him to help cover costs

I'd like your support in getting him to agree to these changes, they would greatly improve the outlook of the project especially since if someone even just wants a small company to sell DRM they can do so without having to worry about high costs. (i want to open one )

Don't mistake me for trying to get in easy, i own a BN that i am running on one of the p2pools (i took over the pools to allow him time to work)

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...