Bitcoin Forum
January 27, 2020, 02:48:49 AM *
News: Latest Bitcoin Core release: 0.19.0.1 [Torrent]
 
   Home   Help Search Login Register More  
Pages: « 1 ... 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 [56] 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 ... 199 »
  Print  
Author Topic: [SKY] Skycoin Launch Announcement  (Read 379756 times)
mladen00
Legendary
*
Offline Offline

Activity: 1890
Merit: 1004


K-ing®


View Profile
July 01, 2014, 06:19:47 AM
 #1101

Development Update:

Skywire Meshnet

The design is amazing. Skywire darknet router may be able to replace BGP and existing corporate VPN solutions. Its reverse comparable with IPv6 (IPv6 can be embedded over Skywire). Its very simple.

I dont want to talk much about this. Just have to finish implementation and get source code out there.

Skycoin Application Framework

Extremely exciting.  We have compiler for compiling Golang down to Javascript.

https://github.com/gopherjs/gopherjs

- Aether apps will be in Golang.
- The Golang will be compiled down to Javascript.
- The Javascript app will be stored in key zero in the Aether store.
- The other keys will contain data driving the application
- The app uses the AngularBinding for DOM manipulation
- Some functions will be exposed via API through DOM (DHT lookup, messaging)

This is a major step towards having iOS/Android style Darknet apps running on Skywire darknet.

Skywire Daemon Progress

Almost working. A few changes left. Changes to service introduction packet format, refactoring, removal of 2,000 lines of redundant code and other changes.

Project Management

Some things can be written down and handed by other developers. They are isolated, independent and easy to describe in detail.

Other things are very complicated design problems, where there is not a correct answer. Very difficult to outsource or delegate. These are problems, where if three developers are working on it, they each have own solution and insist it is the best solution. They cant agree at all. If they compromise the design ends up being worse than any of the individual solutions (design by committee).

We are getting through the last of the difficult problems. These design choices relating to how project components are structured, API interface and messaging formats. This is blocking the work the other developers need to do.  We are nearing end of this and can expand development effort soon.

IPO

We do not need the money from the IPO, so internal pressure to launch coin has been low.  However, we should do that immediately.

The blockchain was running in January and we should have kept it running and just made that the blockchain. Remaining changes affecting blockchain should be finalized and then coin will be "launched". However, much work ahead.


Great news for this great coin.
Can you tell us when the IPO distribution will start?

IPO will be like NEM IPO or like NFD IPO or???

respect 4 devs

IOTA
1580093329
Hero Member
*
Offline Offline

Posts: 1580093329

View Profile Personal Message (Offline)

Ignore
1580093329
Reply with quote  #2

1580093329
Report to moderator
Advertised sites are not endorsed by the Bitcoin Forum. They may be unsafe, untrustworthy, or illegal in your jurisdiction. Advertise here.
John_Paul
Sr. Member
****
Offline Offline

Activity: 269
Merit: 250


View Profile
July 01, 2014, 03:42:31 PM
 #1102

Skycoin seems the next biggest thing for cryptos. Hope it will be launched soon for people to see it more clearly. Interested.
rieneuf
Member
**
Offline Offline

Activity: 68
Merit: 10


View Profile
July 01, 2014, 03:46:23 PM
 #1103

 I don't understand it but I'm sure that's something of real value )!
Eagle_Silver
Newbie
*
Offline Offline

Activity: 15
Merit: 0


View Profile
July 01, 2014, 06:30:38 PM
 #1104

I don't understand it but I'm sure that's something of real value )!

I bet no one really understands it. The devs are drawing a picture for a very complicated system. Maybe, they should complete a simple one first, quickly making something work first then adding more complicated features.
yxxyun
Member
**
Offline Offline

Activity: 100
Merit: 10



View Profile
July 02, 2014, 04:49:12 AM
 #1105

most user will only use skycoin's local web wallet, to make transaction,trade, etc.

somebody will run a Obelisk node, to confirm the transaction, use consensus mechanism. I don't know is there any benefit to run a obelisk node, if you run a btc miner to confirm btc transaction,you can get btc, but you run a rippled node to provide consensus to confirm the ripple transaction,you get nothing.

somebody will run a  meshnet node and will get skycoin. but meshnet don't provide consensus to confirm the transaction, it is just a way to distribute skycoin.
JohnWayne02
Newbie
*
Offline Offline

Activity: 15
Merit: 0


View Profile
July 02, 2014, 11:22:33 AM
 #1106

Meshnet is interesting. I guess with meshnet the cryptos will be free from router attack.
BlackShibe1
Sr. Member
****
Offline Offline

Activity: 260
Merit: 250


View Profile
July 02, 2014, 11:25:38 AM
 #1107

Why is so long
Qora, Simcoin, Crypti are faster

Lisk.
    Develop Decentralized Applications & Sidechains in JavaScript with Lisk!
    Website | Blog | BTT Thread | Chat - Be part of the decentralized application movement!
JFK01
Newbie
*
Offline Offline

Activity: 14
Merit: 0


View Profile
July 02, 2014, 03:06:02 PM
 #1108

Why is so long
Qora, Simcoin, Crypti are faster

Because this might be a generation 2.5 crypto.
LemonAndFries
Sr. Member
****
Offline Offline

Activity: 462
Merit: 250


View Profile
July 02, 2014, 03:54:29 PM
 #1109

well doing a wi-fi sort of encrypted system must be harder to do.

I've been waiting for a long time

but if it does everything the devs have promised, this tech will definitely launch crypto into a whole new phase

this type of over the air tech will make it possible for many people to be reached by crypto
PaulWang
Newbie
*
Offline Offline

Activity: 17
Merit: 0


View Profile
July 02, 2014, 08:53:47 PM
 #1110

This takes so long to launch and I am afraid the clones would come out first.
JessicaMILFson
Full Member
***
Offline Offline

Activity: 182
Merit: 100


View Profile
July 02, 2014, 08:55:37 PM
 #1111

I much rather trust the creator of the code then the clones.
PondSea
Legendary
*
Offline Offline

Activity: 1428
Merit: 1000


View Profile
July 03, 2014, 11:16:03 AM
 #1112

nice





░░░░░░░░░▀▀▀█████████
░░░░░░░░░░░░░░░████████
░░░░▄███████▄░░░░████████
░░░░███████████░░░░██████
░░░▀███████████░░░░████░░
███▄░░░░░░░░░░▀████░░░███░░██
█████▄▄▄▄▄▄▄▄▄▄▄████░░░██░░██
█████████████▄░░████░░░░░
░░█████████████░░█████
░░░░█████████▀░░░██████▌
░░░░░░░▀▀▀▀░░░░▄████████▌
░░░░░░░░░░▄▄▄▄███████
SuperNET.org
..BarterDEX..
.
▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬
DECENTRALIZED CRYPTOCURRENCY EXCHANGE
Developed to Unite Coin Communities | ✔ SECURE ✔ FREE ✔ VISIBILITY ✔ EASY INTEGRATION |

▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬

lovely89
Sr. Member
****
Offline Offline

Activity: 378
Merit: 250


View Profile
July 03, 2014, 11:48:52 AM
 #1113

I think the IPO will like Safecoin, price will be down after it's on exchange, ROI is too low.

Skycoin will have a rolling IPO. The IPO will be very small, then a small number of coins will be released on to an exchange each day. If the price decreases, the number of newly issued coins will be decreased to stabilize price. If the price goes too low, the developers will begin buying back coins.

The majority of distribution will go to developers, people involved in marketing and people buying and operating a mesh network nodes. We will distribute between $100 and $150 dollars per meshnet node deployment. If a user spends $200 to deploy a node and receives $150 in Skycoin, this distribution is unlikely to negatively impact price because of Proof of Burn. Distribution an exchange or airdrop, in contrast is likely to drive down price and should be minimized.

The problem with Maidsafe is that it does not have users and does not have volume. To succeed, Skycoin must have a large and rapidly growing userbase and there must be a reason for people to be buying and using Skycoin.

Coins that no one uses and which have no user base are purely speculative. Many people have made money from speculative investments, but there is no rational basis for the price to increase over time. Dogecoin and Bitcoin should experience appreciation because of increasing adaption and a growing userbase. If the Bitcoin userbase doubles and Bitcoin transaction volumes doubles, we would expect Bitcoin price to approximately double, or at least increase.

It is very difficult to understand why Maidsafe would increase from the IPO, without a growing userbase.

This is Come-from-beyond, the only person we know of who knows who BCNext is and is the main contractor for Nxt. These are his thoughts on 100% POS IPOs.

https://nxtforum.org/initial-distribution/initial-distribution-of-100-pos-currencies/msg35870/#msg35870

Bitrated user: vanlovely.
YNWA2806
Sr. Member
****
Offline Offline

Activity: 400
Merit: 250


View Profile
July 03, 2014, 03:01:15 PM
 #1114


Didn't understand whether this is going to be a standard Pre-Sale or a free distribution....
PeterLong
Newbie
*
Offline Offline

Activity: 15
Merit: 0


View Profile
July 03, 2014, 03:05:13 PM
 #1115

Didn't understand whether this is going to be a standard Pre-Sale or a free distribution....

I think there will be not free coins to distribute. It seems that the issue for now for the dev is how much to sell at the first stage of the IPO or pre sale. It could be from 1% to maybe 10% of total coins.
lcharles123
Legendary
*
Offline Offline

Activity: 1162
Merit: 1013


View Profile
July 03, 2014, 04:49:39 PM
 #1116

Following.  Smiley Translated to portuguese.
JacksonHu
Newbie
*
Offline Offline

Activity: 16
Merit: 0


View Profile
July 07, 2014, 04:22:13 PM
 #1117

How far will it be between the Skycoin and the Skynet?
skycoin
Hero Member
*****
Offline Offline

Activity: 498
Merit: 500


View Profile WWW
July 08, 2014, 04:34:56 AM
 #1118

Development Update

Its July. Meshnet was promised in July. Spec is done. Implementation in progress.

We end up writing everything five times and refactoring it before its perfect. Sorry this takes so long.

Summary of Skycoin Consensus Algorithm Research to Date

We solved adversarial byzantine generals for full connected graphs with a
- cryptographicly constructed "public broadcast channel" with fully connected nodes, but solution does not scale to 300,000 nodes.
-> global consensus must result from local consensus in practical system
-> Proved global consensus from local consensus in random graphs using probabilistic Ben-Ors consensus process

We have mathematical proof that 51% attack proof is achievable with a single trusted timestamping server.
- distributed time stamping did not meet conditions for proof
- However, distributed time stamping useful for detecting 51% attack.
- If 51% attack can be detected reliably, it can be eliminated. Practical 51% attack now require compromosing two seperate indepedent systems.

Mathematical Results:
- There is no deterministic adversarial Paxos solution for the blockchain.
- However, probability of attack success can be reduced to being exponentially small in block confirmation depth using "randomization"
- practical, miner-less 51% attack proof consensus is possible, but require trade offs

Threat Reduction Measures
- increase cost of 51% attack (increase resource cost)
- make 51% attack success probabilistic (decrease payout to attacker)
- allow identification of attacking nodes (auditing, identity)
- when detected, attacking nodes automatically lose resources needed for attack (node social trust metric as scarce resource for network consensus)
- bound time frame over which transactions can be reverted.

Current Skycoin Consensus System Design:
- Public Broadcast Channel for consensus negotiation
- BenOrs randomized consensus protocol in web of trust network
- Detect attack states, mitigate attack, identify attacking nodes
- Multiple layered detection systems that must be defeated for a successful attack.

Consensus

The formal state transition diagram for network consensus for netsplit and 51% attack prevention is coming together. We are sure that if 90% of the nodes in a random graph attempt to revert block consensus, they will fail. There are three difficult conditions that must be met in succession for a block consensus reversion attack to succeed.

There were two states for nodes. A new third state for network bifurcation has been added. This third state simplifies the state transition diagram for dealing with attacks and netsplits
- converged with network (this is determined at minimum by the "Consensus Oracle" mechanism)
- Not converged. This is state where the current node block consensus is on a different chain than rest of network and knows it.
- A new "bifurcated network state". This is the state a node goes into when it detects two or more divergent consensus chains on network subgraphs. This is the state the network is left in after a netsplit or consensus reversion attack attempt. This state begins conflict resolution.

There is a window of blocks. Nodes can remain in the converged state during the consensus process for this many blocks in the presence of blockchain forks while remaining in the converged state.
- This is the "consensus window" in which the network has to resolve choices between alternative blockchains.
- Once a block goes outside of window, it is "final" and reversion should be impossible.
- It is acceptable for the network to stop until consensus is achieved if the window would be exceeded by the additional of new blocks. Stopping or slowing down the network during an attack is preferable to introducing a 51% attack. Once attacking nodes have been identified, people can remove them from their trust lists and their influence on the network eliminated.
- The probability of forks decreases exponentially with block depth within the consensus window

Unconverged states happen when consensus process runs on and exceeds the window. This may indicate a pathological subgraph slowing down convergence or a DDoS attack. It may also reflect lose of connectivity between continents slowing down blockchain convergence.

Note: For the purposes of the finite state machine transition diagram, the unconvergence state is discrete. Block introduction is rejected or discouraged while pending consensus decisions are processed. As a practical matter, there may be transaction throttling, with block rate decreasing as a function of the size of the open decision set. This is called the "soft throttle". The "hard throttle" is when nodes begin rejecting block introduction until the current backlog of consensus decisions are globally resolved.

Bifuricated network state occurs when peers announce a chain which forks before the window and which was not previously in the consensus decision set (fork introduction). This only occurs doing a netsplit when nodes rejoin the network or during a directed attack. Most nodes in the bifurcated network state resolve automatically.

There are several detectable error states, that should not occur in normal operation but will occur. Such as local node assumes a consensus state that differs from the state suggested by the consensus oracle.

Handling Bifurication State:

This state should never be reached, except in emergencies and major coordinated attacks. An attacker must have the resources and the will to attempt to put the network into this state. This will never occur on a network of non-malicious unmodified clients running on a reliable internet.

If this network state has been reached
- There is a severe Skycoin bug that needs to be patched
- Skycoin is under attack from a powerful group with significant resources
- World World III has begin. Global internet connectivity has been severed. The colocation centers have been nuked. The undesea cables have been cut and communication satellites no longer function.

This is the state that triggers the 51% attack detection. This is being worked through. This state should not be triggered during normal operation.

- If someone manages to get enough nodes colluding to influence the consensus, they can DDoS attack the network by minting blocks with no transactions in them and preventing network from accepting valid blocks. This wont last for long. As soon as the nodes are identified, people will remove them from trust lists and its effectively a ban.
- Then there are planned measures in place, that make node consensus decision deterministic. This makes it very difficult for a large number of nodes to collude to influence voting, even if they are controlled by a single person. It turns vote manipulation into an intractable and frustrating optimization and coordination problem.
- If someone tries to 51% attack, it fails with high probability, deterring an attack
- If someone 51% attacks and succeeds, it will trigger automatic resolution, which will block most attacks based upon the block timestamps.
- If they get past the automatic resolution procedures, then it would go into manual resolution.

The network is designed to never get to this point, but if it does then there are a number of possible resolution procedures we are looking at. The network will run both branches concurrently until the split is resolved.
- manual resolution (each user setting chain by hand). The honest nodes and exchanges affected by the split, will just revert the fork and their nodes will discard the bad chain. In Bitcoin the pools participating in an attack are likely to be in on the attack and the victims are the exchanges and service operators, making it difficult to revert the chain after an attack. In Skycoin, people are unlikely to use a chain that is not the consensus chain of the large exchanges and Skycoin service providers (who are the victims and targets of the double spending attack). The incentives are therefore slightly better aligned.
- A developer emergency signing key with 2 out of 3 keys required. Instantly resolves deadlock. Should never get to this point.

Transaction Confirmation States and Block Depth:

There is a separate flowchart for transaction confirmation. Transactions will be reported by the client JSON API as
- pending: transaction relayed by network, but not entered into block yet
- confirmed: transaction entered into block in consensus chain with N confirmations.

N confirmations means the block in which the transaction was entered, has N successor blocks up to the first block which has 0 or more than one remaining candidate child blocks pending network consensus. Skycoin can have multiple child blocks and branches pending consensus, the network does not slow down to decide a single block and then the network block. Therefore each transaction has a "depth" with respect to the last block with more than two remaining parents and the transaction has a depth with respect to each of the pending consensus chains.

So for example, if a netsplit occurs at block 10. The netsplit chain is on block 20. A transaction is entered in netsplit chain in block 15. The transaction has five confirmations with respect to the head of the netsplit chain, but has 0 confirmations (pending state) because block 10 has two children (block 10 was last block when the local node was in the "converged" state).
- If the netsplit chain and mainchain merge at block 25, if the netsplit chain becomes dominant and the other chain pruned, then the transaction now has 10 confirmations
- The transaction can never go into the confirmed stated while the node is in unconverged or bifurcated state. This automatically prevents the vast majority of double spending attacks against exchanges and merchants, as most exchanges will not apply a credit to an account until the transaction has reached N confirmations.
- In Bitcoin "Confirmations" are the number of blocks since the block containing the transaction, for the longest block. In Skycoin, confirmations are the number of blocks after the transaction which have

There is a closely related concept, the "confirmation min" and the "confirmation max". If there are two chain branches open for consensus and the transaction has four confirmations on the first chain and six confirmations on the second chain, the transaction will have a minimum of 4 confirmations regardless of which chain is chosen for consensus.

Scenario
- there are two blockchain branches
- The first branch has 15 blocks
- The second branch as 12 blocks
- The branches fork at block 7 (block seven is the last block the two branches have in common)
- A transaction was entered into the first branch at block 11 and the second branch at block 10
Therefore
- the transaction has "0 confirmations" because the transaction was not entered before the chain forks on block 7
- the transaction has 4 confirmations on the first branch
- the transaction has 2 confirmations on the second branch
- regardless of which branch is chosen for consensus, the transaction will be executed (unless one forks is chosen and then another fork occurs at a block the transaction was executed in)

There is a calculus of blockchains. There is a min, a max, a join and a bottom. We need to develop good terminology for talking about the blockchain forks, diagrams and software examples.  Documentation is low priority until its working and the wiki is up. Someone will have to make a tutorial on the wiki, make python scripts for generating diagrams and come up with terminology.

HjalmarX
Newbie
*
Offline Offline

Activity: 41
Merit: 0


View Profile
July 08, 2014, 04:38:47 AM
 #1119

what is outlook now for next two weeks.

what needs to be done, what will happen?
FrictionlessCoin
Legendary
*
Offline Offline

Activity: 868
Merit: 1000


Cryptotalk.org - Get paid for every post!


View Profile
July 08, 2014, 01:06:53 PM
 #1120

Development Update

Its July. Meshnet was promised in July. Spec is done. Implementation in progress.

We end up writing everything five times and refactoring it before its perfect. Sorry this takes so long.

Summary of Skycoin Consensus Algorithm Research to Date

We solved adversarial byzantine generals for full connected graphs with a
- cryptographicly constructed "public broadcast channel" with fully connected nodes, but solution does not scale to 300,000 nodes.
-> global consensus must result from local consensus in practical system
-> Proved global consensus from local consensus in random graphs using probabilistic Ben-Ors consensus process

We have mathematical proof that 51% attack proof is achievable with a single trusted timestamping server.
- distributed time stamping did not meet conditions for proof
- However, distributed time stamping useful for detecting 51% attack.
- If 51% attack can be detected reliably, it can be eliminated. Practical 51% attack now require compromosing two seperate indepedent systems.

Mathematical Results:
- There is no deterministic adversarial Paxos solution for the blockchain.
- However, probability of attack success can be reduced to being exponentially small in block confirmation depth using "randomization"
- practical, miner-less 51% attack proof consensus is possible, but require trade offs

Threat Reduction Measures
- increase cost of 51% attack (increase resource cost)
- make 51% attack success probabilistic (decrease payout to attacker)
- allow identification of attacking nodes (auditing, identity)
- when detected, attacking nodes automatically lose resources needed for attack (node social trust metric as scarce resource for network consensus)
- bound time frame over which transactions can be reverted.

Current Skycoin Consensus System Design:
- Public Broadcast Channel for consensus negotiation
- BenOrs randomized consensus protocol in web of trust network
- Detect attack states, mitigate attack, identify attacking nodes
- Multiple layered detection systems that must be defeated for a successful attack.

Consensus

The formal state transition diagram for network consensus for netsplit and 51% attack prevention is coming together. We are sure that if 90% of the nodes in a random graph attempt to revert block consensus, they will fail. There are three difficult conditions that must be met in succession for a block consensus reversion attack to succeed.

There were two states for nodes. A new third state for network bifurcation has been added. This third state simplifies the state transition diagram for dealing with attacks and netsplits
- converged with network (this is determined at minimum by the "Consensus Oracle" mechanism)
- Not converged. This is state where the current node block consensus is on a different chain than rest of network and knows it.
- A new "bifurcated network state". This is the state a node goes into when it detects two or more divergent consensus chains on network subgraphs. This is the state the network is left in after a netsplit or consensus reversion attack attempt. This state begins conflict resolution.

There is a window of blocks. Nodes can remain in the converged state during the consensus process for this many blocks in the presence of blockchain forks while remaining in the converged state.
- This is the "consensus window" in which the network has to resolve choices between alternative blockchains.
- Once a block goes outside of window, it is "final" and reversion should be impossible.
- It is acceptable for the network to stop until consensus is achieved if the window would be exceeded by the additional of new blocks. Stopping or slowing down the network during an attack is preferable to introducing a 51% attack. Once attacking nodes have been identified, people can remove them from their trust lists and their influence on the network eliminated.
- The probability of forks decreases exponentially with block depth within the consensus window

Unconverged states happen when consensus process runs on and exceeds the window. This may indicate a pathological subgraph slowing down convergence or a DDoS attack. It may also reflect lose of connectivity between continents slowing down blockchain convergence.

Note: For the purposes of the finite state machine transition diagram, the unconvergence state is discrete. Block introduction is rejected or discouraged while pending consensus decisions are processed. As a practical matter, there may be transaction throttling, with block rate decreasing as a function of the size of the open decision set. This is called the "soft throttle". The "hard throttle" is when nodes begin rejecting block introduction until the current backlog of consensus decisions are globally resolved.

Bifuricated network state occurs when peers announce a chain which forks before the window and which was not previously in the consensus decision set (fork introduction). This only occurs doing a netsplit when nodes rejoin the network or during a directed attack. Most nodes in the bifurcated network state resolve automatically.

There are several detectable error states, that should not occur in normal operation but will occur. Such as local node assumes a consensus state that differs from the state suggested by the consensus oracle.

Handling Bifurication State:

This state should never be reached, except in emergencies and major coordinated attacks. An attacker must have the resources and the will to attempt to put the network into this state. This will never occur on a network of non-malicious unmodified clients running on a reliable internet.

If this network state has been reached
- There is a severe Skycoin bug that needs to be patched
- Skycoin is under attack from a powerful group with significant resources
- World World III has begin. Global internet connectivity has been severed. The colocation centers have been nuked. The undesea cables have been cut and communication satellites no longer function.

This is the state that triggers the 51% attack detection. This is being worked through. This state should not be triggered during normal operation.

- If someone manages to get enough nodes colluding to influence the consensus, they can DDoS attack the network by minting blocks with no transactions in them and preventing network from accepting valid blocks. This wont last for long. As soon as the nodes are identified, people will remove them from trust lists and its effectively a ban.
- Then there are planned measures in place, that make node consensus decision deterministic. This makes it very difficult for a large number of nodes to collude to influence voting, even if they are controlled by a single person. It turns vote manipulation into an intractable and frustrating optimization and coordination problem.
- If someone tries to 51% attack, it fails with high probability, deterring an attack
- If someone 51% attacks and succeeds, it will trigger automatic resolution, which will block most attacks based upon the block timestamps.
- If they get past the automatic resolution procedures, then it would go into manual resolution.

The network is designed to never get to this point, but if it does then there are a number of possible resolution procedures we are looking at. The network will run both branches concurrently until the split is resolved.
- manual resolution (each user setting chain by hand). The honest nodes and exchanges affected by the split, will just revert the fork and their nodes will discard the bad chain. In Bitcoin the pools participating in an attack are likely to be in on the attack and the victims are the exchanges and service operators, making it difficult to revert the chain after an attack. In Skycoin, people are unlikely to use a chain that is not the consensus chain of the large exchanges and Skycoin service providers (who are the victims and targets of the double spending attack). The incentives are therefore slightly better aligned.
- A developer emergency signing key with 2 out of 3 keys required. Instantly resolves deadlock. Should never get to this point.

Transaction Confirmation States and Block Depth:

There is a separate flowchart for transaction confirmation. Transactions will be reported by the client JSON API as
- pending: transaction relayed by network, but not entered into block yet
- confirmed: transaction entered into block in consensus chain with N confirmations.

N confirmations means the block in which the transaction was entered, has N successor blocks up to the first block which has 0 or more than one remaining candidate child blocks pending network consensus. Skycoin can have multiple child blocks and branches pending consensus, the network does not slow down to decide a single block and then the network block. Therefore each transaction has a "depth" with respect to the last block with more than two remaining parents and the transaction has a depth with respect to each of the pending consensus chains.

So for example, if a netsplit occurs at block 10. The netsplit chain is on block 20. A transaction is entered in netsplit chain in block 15. The transaction has five confirmations with respect to the head of the netsplit chain, but has 0 confirmations (pending state) because block 10 has two children (block 10 was last block when the local node was in the "converged" state).
- If the netsplit chain and mainchain merge at block 25, if the netsplit chain becomes dominant and the other chain pruned, then the transaction now has 10 confirmations
- The transaction can never go into the confirmed stated while the node is in unconverged or bifurcated state. This automatically prevents the vast majority of double spending attacks against exchanges and merchants, as most exchanges will not apply a credit to an account until the transaction has reached N confirmations.
- In Bitcoin "Confirmations" are the number of blocks since the block containing the transaction, for the longest block. In Skycoin, confirmations are the number of blocks after the transaction which have

There is a closely related concept, the "confirmation min" and the "confirmation max". If there are two chain branches open for consensus and the transaction has four confirmations on the first chain and six confirmations on the second chain, the transaction will have a minimum of 4 confirmations regardless of which chain is chosen for consensus.

Scenario
- there are two blockchain branches
- The first branch has 15 blocks
- The second branch as 12 blocks
- The branches fork at block 7 (block seven is the last block the two branches have in common)
- A transaction was entered into the first branch at block 11 and the second branch at block 10
Therefore
- the transaction has "0 confirmations" because the transaction was not entered before the chain forks on block 7
- the transaction has 4 confirmations on the first branch
- the transaction has 2 confirmations on the second branch
- regardless of which branch is chosen for consensus, the transaction will be executed (unless one forks is chosen and then another fork occurs at a block the transaction was executed in)

There is a calculus of blockchains. There is a min, a max, a join and a bottom. We need to develop good terminology for talking about the blockchain forks, diagrams and software examples.  Documentation is low priority until its working and the wiki is up. Someone will have to make a tutorial on the wiki, make python scripts for generating diagrams and come up with terminology.

This is all great stuff.  Kudos on the great work.   I rather that you folks get it right rather than release a half-baked coin like so many others have done.

 
                                . ██████████.
                              .████████████████.
                           .██████████████████████.
                        -█████████████████████████████
                     .██████████████████████████████████.
                  -█████████████████████████████████████████
               -███████████████████████████████████████████████
           .-█████████████████████████████████████████████████████.
        .████████████████████████████████████████████████████████████
       .██████████████████████████████████████████████████████████████.
       .██████████████████████████████████████████████████████████████.
       ..████████████████████████████████████████████████████████████..
       .   .██████████████████████████████████████████████████████.
       .      .████████████████████████████████████████████████.

       .       .██████████████████████████████████████████████
       .    ██████████████████████████████████████████████████████
       .█████████████████████████████████████████████████████████████.
        .███████████████████████████████████████████████████████████
           .█████████████████████████████████████████████████████
              .████████████████████████████████████████████████
                   ████████████████████████████████████████
                      ██████████████████████████████████
                          ██████████████████████████
                             ████████████████████
                               ████████████████
                                   █████████
.CryptoTalk.org.|.MAKE POSTS AND EARN BTC!.🏆
Pages: « 1 ... 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 [56] 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 ... 199 »
  Print  
 
Jump to:  

Sponsored by , a Bitcoin-accepting VPN.
Powered by MySQL Powered by PHP Powered by SMF 1.1.19 | SMF © 2006-2009, Simple Machines Valid XHTML 1.0! Valid CSS!