elective-stereophonic
elective-stereophonic
Show Posts - andruiman singapore
Please login or register.

Login with username, password and session length
Advanced search  

News:

Latest Nxt Client: Nxt 1.11.15

Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.

Messages - andruiman

Pages: [1] 2
1
Nxt Asset Exchange / Re: LISK 1:1 Backed Asset
« on: June 07, 2016, 04:23:10 pm »
To the lucky person who bought please send the assets back to me with a message attached stating your desired LISK address you would like the LISK to be sent to. The message should be public.

Thank you!

Hi, I've sent bought assets back (txid: 4807636929649032977) some time ago, could you please let me know if it's OK to redeem them to the LISK tokens (the address is specified in the message). Thanks.

2
SuperNET Releases / Re: Multigateway User Support Thread
« on: October 26, 2015, 06:05:47 am »
Hi

I have missed the mgwBTC -> superBTC migration, so I have occasionally sent some coins to the mgwBTC deposit address.

txid is 4716f3a698986d4e2d46e8222d534df60976a64dc7e622403e0c1b32086aa6e6.

Are the coins recoverable?

Thanks.

Yes, there should be no problem. There's still some late deposits received in old wallets so we need to manually transfer funds to new servers and then send superBTC to the depositors. Please send me the Nxt account where you want the assets delivered and allow for a few days.

Thank you. The Nxt account is NXT-L892-ZKXZ-2JJY-AD9JV.

Everything arrived. Thanks.

3
SuperNET Releases / Re: Multigateway User Support Thread
« on: October 24, 2015, 01:06:23 am »
Hi

I have missed the mgwBTC -> superBTC migration, so I have occasionally sent some coins to the mgwBTC deposit address.

txid is 4716f3a698986d4e2d46e8222d534df60976a64dc7e622403e0c1b32086aa6e6.

Are the coins recoverable?

Thanks.

Yes, there should be no problem. There's still some late deposits received in old wallets so we need to manually transfer funds to new servers and then send superBTC to the depositors. Please send me the Nxt account where you want the assets delivered and allow for a few days.

Thank you. The Nxt account is NXT-L892-ZKXZ-2JJY-AD9JV.

4
SuperNET Releases / Re: Multigateway User Support Thread
« on: October 23, 2015, 03:40:54 pm »
Hi

I have missed the mgwBTC -> superBTC migration, so I have occasionally sent some coins to the mgwBTC deposit address.

txid is 4716f3a698986d4e2d46e8222d534df60976a64dc7e622403e0c1b32086aa6e6.

Are the coins recoverable?

Thanks.

5
SuperNET Releases / Re: Multigateway User Support Thread
« on: July 10, 2015, 08:38:07 pm »
Hi

I've made a withdrawal mgwBTC -> BTC (txid 149471786984067940) about 10 hours ago and still have no coins arrived.
Is it typical and I shouldn't worry about?

Thanks

Alright, they are now coming :)
Thanks 

6
SuperNET Releases / Re: Multigateway User Support Thread
« on: July 10, 2015, 07:02:33 pm »
Hi

I've made a withdrawal mgwBTC -> BTC (txid 149471786984067940) about 10 hours ago and still have no coins arrived.
Is it typical and I shouldn't worry about?

Thanks

7
Hi

Could you please describe the current status.
Is the project still alive?
Do you still buy the assets back?

9
Consensus Research / Re: Consensus Research 2015
« on: February 04, 2015, 02:52:02 pm »
We have updated our github repository https://github.com/ConsensusResearch/ForgingSimulation with a new version of the PoS simulation haskell code.
It now included two branches, master - for the single branch classical Nxt based code and "multibranch-experimental" - for the multibranch forging simulation. Recently
new algorithm for regulating tails switching effect is proposed and implemented. With it, a possibility of the N@S attack becomes also regulated as we now can introduce deducible  parameter of confirmations needed to stabilize recent blocks tails. The idea of regulating is straightforward - from time to time the node "forgets" almost all the branches and prolong only those whose cumulativeDifficulty measure is above some retargeting threshold. This threshold changes discretely, starting from 0. Unlike the Bitcoin difficulty param, the threshold always grows as the best block cumulativeDifficulty exceeds the previous threshold+delta. So nodes work as multibranch almost all the time, but sometimes becomes "single-branch" for a short time (one tick). This approach allows to have all the multibranch benefits and also get the network with regulating convergence. With a certain confirmation number  calculated, we can propose the strong resistance to the N@S as the long tails switching become very-very unlikely after the confirmations. We'll present the N@S simulation results ASAP.

There are more possible regulation procedures, for sure. Basing on the idea that sometimes nodes switch to the single-branch behavior one can introduce any verifiable quasi-random algorithm to do this. The proposed is the simple but efficient one, however more complicated algos (e.g. based on some nice hashes) could  secure the system more likely.

New paper on tails switching effect had been publicly released (https://github.com/ConsensusResearch/articles-papers/tree/master/switching). However the results of the simulations presented in the paper have been already renewed by the simulation software at https://github.com/ConsensusResearch/ForgingSimulation/tree/multibranch-experimental with the proposed threshold algorithm. As expected the algorithm allows to have confirmation number parameter deducible from the system constants and prevents the prolongation of similar branches. With it the resistance to the N@S becomes feasible and measurable! The results of N@S simulation + switching tails length distribution are coming.

10
Consensus Research / Re: Consensus Research 2015
« on: January 23, 2015, 02:44:20 pm »
One of the next step of investigation I propose strict formalization of the PoS systems with algebraic approach and Coq specification. It can be considered as a standard for such systems and even published in IEEE or whatever. So I'd like to ask the community to share opinions of such initiative. So the idea is to create a triangle: standard (i.e. spec) -> reference implementation -> production implementations. Nxt IMO is the best platform to start as it has very open community.

(Also the most detailed description of the different measures influence on the multibranch forging is on the way. That will be given to show some difficulties to prevent the branching behavior to diverge - e.g. it cannot be predictable for one branch to be better than the current best in some steps! and as far as I understand at the moment the only way to stop the switching is to stop branching  :)  for a short while).

11
Consensus Research / Re: Consensus Research 2015
« on: January 23, 2015, 02:28:03 pm »
We have updated our github repository https://github.com/ConsensusResearch/ForgingSimulation with a new version of the PoS simulation haskell code.
It now included two branches, master - for the single branch classical Nxt based code and "multibranch-experimental" - for the multibranch forging simulation. Recently
new algorithm for regulating tails switching effect is proposed and implemented. With it, a possibility of the N@S attack becomes also regulated as we now can introduce deducible  parameter of confirmations needed to stabilize recent blocks tails. The idea of regulating is straightforward - from time to time the node "forgets" almost all the branches and prolong only those whose cumulativeDifficulty measure is above some retargeting threshold. This threshold changes discretely, starting from 0. Unlike the Bitcoin difficulty param, the threshold always grows as the best block cumulativeDifficulty exceeds the previous threshold+delta. So nodes work as multibranch almost all the time, but sometimes becomes "single-branch" for a short time (one tick). This approach allows to have all the multibranch benefits and also get the network with regulating convergence. With a certain confirmation number  calculated, we can propose the strong resistance to the N@S as the long tails switching become very-very unlikely after the confirmations. We'll present the N@S simulation results ASAP.

There are more possible regulation procedures, for sure. Basing on the idea that sometimes nodes switch to the single-branch behavior one can introduce any verifiable quasi-random algorithm to do this. The proposed is the simple but efficient one, however more complicated algos (e.g. based on some nice hashes) could  secure the system more likely.

12
"We showed that in comparison with single-branch strategy the multibranch one is more efficient in terms of the number of generated block (and therefore the fee rewarded)"
- Why should this be true? The fees are not dependent on the number of generated blocks but on the number of transactions that are included in the blocks. A transaction usually has a cause and that is more or less given and independent on the way blockchain (or "blocktree") works. Faster generation of blocks only potentially increases the velocity of money but I'm not confident that it would really make a difference.

You are right. But that statement is given about the comparison with single-branch forging, i.e. multibranch account can effectively forge more blocks and get proportionally more fee than a single-branch within the same time interval and ceteris paribus.

13
I don't get how multiple branches can carry the same information. Even if all the branches included the same transactions and malicious attacks were impossible, there still will be differences in the accounts that forged particular blocks and e.g. in the amounts of money on these accounts (because of the fees). What do I miss? A single public ledger (be it blockchain or "blocktree") must always carry single, common information, otherwise it's worthless.

We offer to distinguish the common data and common vision. Generally speaking it is not necessary for accounts to keep the same data to act equally (or equivalently). The main purpose of the blocktree structure is to implement what actually exists. The current Nxt protocol can deal with branches by rolling back blocks, we offer to do it in more natural way keeping the whole tree structure.  There is an inspiring analogue from physics - where are observable and unobservable values, e.g. quantum psi-function with unobservable argument for exp, or vector magnetic potential - all those example leads to very powerful calibration theory and are fundamental property of the nature (as we believe :). In economics this principle also can be applied as we know from the Quantum Finance studies. I spent some of my life to the intellectual property appraisal (math methods) and ready to conclude that the monetary value  of the objects is not observable unless we do a particular transaction, which we can treat as a quantum measurement. The great example could be also taken from the robotics where the pioneering works of S. Thrun gave the probabilistic approach to the self-localization problem as opposite to deterministic.

The same principles we are implementing here through the blocktree structure giving the possibility to maintain maybe different underlying data with the equivalent reasoning.   

14
Consensus Research / Re: Multibranch forging approach
« on: December 20, 2014, 10:33:15 pm »
Coq sources have been published! link removed

Thanks for doing and publishing these Coq formulations.

I'm having a little problem with POStructures.v:

coqc POStructures.v
...
File "./POStructures.v", line 5, characters 0-25:
Error: Cannot find library FixedList in loadpath

I didn't find FixedList in the Coq standard library. Do you know where I can find it?

Sorry, actually this dependence is unnecessary now. Updated POStructures.v is committed to the repository.
Also I have temporary removed the POSimul5.v as it is a little incompatible with the current version of POStructures.
I will add it in a couple of days.

15
Consensus Research / Re: The Paper on Long-Range attack & Nothing-at-Stake
« on: December 18, 2014, 03:06:50 pm »
Yes, that is my understanding as well. As explained in the post above, I think long-range attack is impossible because entropy change would be too high. It is a sane default for a blockchain system to impose some rules whether to chain "re-writes". However such rules should be algorithmic. I'd add that in reality there is no such thing as "hidden" chain. There is longest chain which has to be accepted by default through the algorithm. Crafting of a chain is not really possible (nobody can make up balances just like that). So the N@S descriptions have been in wrong with regards to these problems, IMO. Results of this paper are roughly the similar I think.
Maybe I do not get the message fully, so sorry in this case.

From the current vision I do not see any connection between long range attack and N@S in the formulations given in the paper. We define the long range attack as a possibility to produce a better chain alone by some account (well, it can contain the same transactions and other stuff, so the balance entropy approach  can pass it). The result, attacker can potentially get, is e.g. about the rewarded fee or slight manipulation of transactions - not critical. However from our investigation the long range attack is quite difficult to execute.

The N@S attack in our formulation is about double spending - nothing hidden at all. The concurrent chains are almost the same and I do not see any statistical measures to distinguish between them. The only difference between them is two transactions included separately.  The solution could lie in algorithmically defined and community accepted blockchain measure function which allows not to jump from the confirmed(!) chain to another.  Unlikely it can be based on the statistical properties.

16
Consensus Research / Re: The Paper on Long-Range attack & Nothing-at-Stake
« on: December 18, 2014, 02:21:45 pm »
I'm thinking specifically of an attack whereby an attacker tracks down a number of the people who were part of the initial genesis block of Nxt. They manage to buy the addresses used, now empty of current Nxt and worthless to the owners, and manage to hold the addresses of a large amount of historical Nxt. (Say 70%+). They start forging a new chain using this 70%, fastwording through blocks, faking checkpoints and maybe faking transaction volume until this new chain catches up with real chain. Now, to someone already on the main chain, they aren't going to mistake this new chain for the real chain. But when someone joins the network for the first time, could they be fooled into thinking that the new chain is the real one?

Yes, I think that in the pure abstract system (without some magic hardcoded, nodes ratings etc) it could be possible to find a better chain doing multibranching with 70% of stake (or even less) than ~100% singlebranching. However IMO it is quite easy to resist it by the concrete design.

17
Consensus Research / Re: The Paper on Long-Range attack & Nothing-at-Stake
« on: December 18, 2014, 01:29:01 pm »
Exciting stuff, great to see such in-depth analysis of PoS. Would love to get input from such guys as Vitalik Buterin and Peter Todd and PoW specialists. Especially if you figure out the "solution to make consensus algorithmically enforced". Be good to get PoS specialists like Sunny King and Rat4 involved in the conversation too.

The paper is really well written and surprisingly understandable. (surprising because it so also very technical)

Did you address what Buterin calls weak subjectivity? The idea that if someone joins the network cold, they need to be told the branch to forge on. If someone got the keys to empty genesis accounts with a large % of forging power, could that person forge an alternative chain and fool those who enter the network cold?

You didn't mention checkpoints either, are they not important under this model?

Thank you! If I understand you correctly that is the same we called "Hidden multibranch" attack, so we think that it is practically impossible to build a chain better than the rest of the network. The key things are the measure function and baseTarget validation procedure. So alone forger should have very significant stake (around 25%+) to produce a better chain. However a short range attacks are still possible due to the block delays in the "best" chain if the rest of the network remains singlebranch and the attacker does multibranching. The easiest opposition is to be multibranch for the network itself.

We do not use checkpoints in the model as we hope the solution could be achieved using the correct block measure function. I suggest however that in practice they could be used to stabilize the branching for hard cases.

18
Consensus Research / Re: The Paper on Long-Range attack & Nothing-at-Stake
« on: December 18, 2014, 01:08:47 pm »
Fantastic!
Consensus = Best Asset :D

Do I get the following right (since I want to add citations to the NXT wikipedia article)?
This is about transparent forging - which is called multibranch since it's a specific implementation. As such the paper defines N@S but the security results are mostly about transparent forging so don't apply to the current NXT blockchain.
You refer to the current chain as "Single-branch". And it is in fact more secure vs. N@S?

Thank you! Actually the current implementation of Nxt is single-branch however the principle of the network allows to be the multibranch, so it is not hardcoded to be singlebranch for a arbitrary participant. So the security of the Nxt is the same as for its multibranch implementation, even more - from our current vision the security of total multibranch network is higher than for its singlebranch approximation. N@S can be more easily executed for the system with low level of branching (cause it's cheaper) than for the system with high level of  branching.

As for TF - multibranching is definitely a kind of implementation, but we see there additional important properties which allow to stabilize the block rate and effectively resist to a number of attacks. 

19
Consensus Research / Re: Multibranch forging approach
« on: November 27, 2014, 07:21:40 am »
4. We can put Latex sources in repo if that matters

done

20
Consensus Research / Re: Multibranch forging approach
« on: November 26, 2014, 11:28:23 am »
ELI5, multibranching is similar to TF with forgers who try to increase number of forged blocks?

Yes, actually there could be applied different measures on branches to choose the best. In our simulation we use the unit measure that is cumulative weight
of a chain is equal to its length. So we separate the forging on two phases: (1) building blocks and (2) choosing a branch. When building blocks nodes share them in a tree-like way and that is "transparency", when choosing the branch they use their strategy which generally speaking can be selected independently.

Pages: [1] 2
elective-stereophonic
elective-stereophonic
assembly
assembly