Cleanup: I'll attack some coins - I owned APEXcoin for 90 blocks

[fenghush]

But killing zombies is a lot of fun.

yay we have the right to destroy peoples work, decide which coins are "shitcoins" and which ones are "good", which coins have potential or should just be killed... cause we're a bunch of   https://www.youtube.com/watch?v=UrgpZ0fUixs


notice how most true shitcoins die off on their own

agreed and +1
At the end of 2015 many coins will be dead. (Without any attack)

But I want my coin secure. The possibility of an attack is reason enough for me.
Shitcoins will not update their code. Most of them have no active development.
Like I said: Attacking a shitcoin is like killing a zombie. Nobody cares. Look at Apex.

Cheers,
Ray

[1715110399]

1715110399

[1715110399]

1715110399

[1715110399]

1715110399

[1715110399]

1715110399

[1715110399]

1715110399

[1715110399]

1715110399

[Rent_a_Ray]

Do you really want to kill that beautiful lady?

[cynicSOB]

But killing zombies is a lot of fun.

so is this

c'mon, APEX's price hasn't moved, and they didn't even have checkpointing servers (not that it would have helped, but It shows they didn't care or they were not prepared to run a coin).
I like to think I'm not destroying work, I'm building a more secure future. For our kids! would anybody think about the children !?

[fonzerrellie]

But killing zombies is a lot of fun.

yay we have the right to destroy peoples work, decide which coins are "shitcoins" and which ones are "good", which coins have potential or should just be killed... cause we're a bunch of   https://www.youtube.com/watch?v=UrgpZ0fUixs


notice how most true shitcoins die off on their own

agreed and +1
At the end of 2015 many coins will be dead. (Without any attack)

But I want my coin secure. The possibility of an attack is reason enough for me.
Shitcoins will not update their code. Most of them have no active development.
Like I said: Attacking a shitcoin is like killing a zombie. Nobody cares. Look at Apex.

Cheers,
Ray

exactly killing next to dead coins isn't impressive, nor does it really show what would happen to a "good" coin with a large network. wouldn't using testnet be much more useful?  

I never really followed Apex, was it yet another one with a ditching dev like mobcoin, nebula, xanon, etc, etc , etc ,etc ? cause non of them are on exchanges or have people trying to make something happen with them. (even if their just bag holders looking for a dev to take over)  just dead

apex was still on an exchange even if any trading movement was just to get the .2 btc daily so they didn't get kicked off the exchange.

I don't know seems like if you guys were just interested in making good coins safer you'd use testnet, find weaknesses and make coins stronger... if your killing zombies just for fun... your still killing just for fun and it's still quite psychotic

[Rent_a_Ray]

But killing zombies is a lot of fun.

so is this

c'mon, APEX's price hasn't moved, and they didn't even have checkpointing servers (not that it would have helped, but It shows they didn't care or they were not prepared to run a coin).
I like to think I'm not destroying work, I'm building a more secure future. For our kids! would anybody think about the children !?

Hey cynicSOB,
what do you think about that: https://bitcointalk.org/index.php?topic=897493.msg10208018#msg10208018

Would it lower your success?

Cheers,
Ray

[cynicSOB]

But killing zombies is a lot of fun.

so is this

c'mon, APEX's price hasn't moved, and they didn't even have checkpointing servers (not that it would have helped, but It shows they didn't care or they were not prepared to run a coin).
I like to think I'm not destroying work, I'm building a more secure future. For our kids! would anybody think about the children !?

Hey cynicSOB,
what do you think about that: https://bitcointalk.org/index.php?topic=897493.msg10208018#msg10208018

Would it lower your success?

Cheers,
Ray

I think it wouldn't work: how can you tell an attack from normal use? if you have 2 chains that look similar but have one conflicting tx you have to choose one, there's just no way to avoid that. How do you define which is the "good" one? an attack may not even change the difficulty

[cynicSOB]

I don't know seems like if you guys were just interested in making good coins safer you'd use testnet, find weaknesses and make coins stronger...

testnet is not enough for some security research
the blockchain doesn't behave exactly the same, and the coin being half-alive and in an exchange gives me the opportunity to see how they react to the attack

[Este Nuno]

But killing zombies is a lot of fun.

yay we have the right to destroy peoples work, decide which coins are "shitcoins" and which ones are "good", which coins have potential or should just be killed... cause we're a bunch of   https://www.youtube.com/watch?v=UrgpZ0fUixs


notice how most true shitcoins die off on their own

agreed and +1
At the end of 2015 many coins will be dead. (Without any attack)

But I want my coin secure. The possibility of an attack is reason enough for me.
Shitcoins will not update their code. Most of them have no active development.
Like I said: Attacking a shitcoin is like killing a zombie. Nobody cares. Look at Apex.

Cheers,
Ray

exactly killing next to dead coins isn't impressive, nor does it really show what would happen to a "good" coin with a large network. wouldn't using testnet be much more useful? 

I never really followed Apex, was it yet another one with a ditching dev like mobcoin, nebula, xanon, etc, etc , etc ,etc ? cause non of them are on exchanges or have people trying to make something happen with them. (even if their just bag holders looking for a dev to take over)  just dead

apex was still on an exchange even if any trading movement was just to get the .2 btc daily so they didn't get kicked off the exchange.

I don't know seems like if you guys were just interested in making good coins safer you'd use testnet, find weaknesses and make coins stronger... if your killing zombies just for fun... your still killing just for fun and it's still quite psychotic

He is using testnet as well. In fact he's just been working on NXT testnet at the request of many. I'm guessing you haven't read the thread thus far. He's doing a good job so far and lots of people are interested in seeing more of his attempts.

This type of thing is critical for PoS to really be considered a robust consensus system. Actually it's not great that it's taken this long for some good public stress testing/attacks. Anyone interested in the idea of PoS should be supporting people who can help strengthen it.

[Rent_a_Ray]

My thought was, that the high diff comes from one single node.
That could be a major stakeholder or an attacker. But I am with you. Idea has gaps.

Whatever, someone will find a solution.  

Cheers,
Ray

[illodin]

cynicSOB, what would be the required countermeasure for a staking coin? decrease the maximum coin age?

https://blackcoin.co/blackcoin-pos-protocol-v2-whitepaper.pdf

[Daedelus]

cynicSOB, what would be the required countermeasure for a staking coin? decrease the maximum coin age?

https://blackcoin.co/blackcoin-pos-protocol-v2-whitepaper.pdf

These changes made BC less like Peercoin and more like Nxt (possibly, exactly the same as Nxt. What are the chances.. ) .

Improvements on their proof of stake algorithm by blackcoin: http://www.blackcoin.co/blackcoin-pos-protocol-v2-whitepaper.pdf Any relevance to Nxt?

They removed "age" and their PoS is very similar* to Nxt PoS now.

-------------
* - I would say "exactly the same" but some details r not clear.

https://nxtforum.org/general-discussion/price-speculation/msg56825/#msg56825

That only Blackcoin developers made security patches and changes to "POS System (PoS2.0)" to fix problems.

The funny thing is that Blackcoin changes in PoS system led to the scheme that has been used in Nxt since day 1.  

https://nxtforum.org/general-discussion/price-speculation/msg56825/#msg56825

[Brangdon]

I'll tell you my theories anyway: the two PhDs who wrote that research have a clear nxt bias. Look at their hidden multibranch section: they say that they got 3/20 of 500 blocks with 10% stake. 3/20 of 500 is 75 blocks. They got 75 blocks! if people wait for 6 confirmations, I only need 6 blocks, not 75. So I can do with much less than 10% stake. Of course 10% stake means 10% of actively forging coins, which means 10% of 10% of supply. So, much less than 1% of supply to double spend. Of course that's still a few hundred BTCs for NXT so I'll start with a smaller coin.

I already bought a few coins... just wait until age is accumulated...

Nxt recommends 10 blocks for normal confirmation, 720 to be sure. (There's a rolling check point at 720.)

Generally Nxt has 40-50% forging, so 10% stake means 4-5% of supply.

Sure, thanks. But attacking nxt requires funding that I don't have. I still claim that it can be attacked with less than 10%
NXT may not have age since last transaction, but it has "age" since last block. It accumulates equally to all miners ("forgers"?), but I believe it can be gamed to obtain more than the fair share of probabilities of creating a fork.

I don't know what you're talking about here. Nxt makes newly moved coins wait 1440 blocks before their weight is counted, but after that their weight counts in every block. I don't know what you mean by "age [that] accumulates equally to all miners". The chance of forging a block does not depend on how long it is since the previous block you forged. It depends on what fraction of the forging coins you have.

What is your criteria for a successful attack?

[TwinWinNerD]

I like how the NXT community is pushing for an attack on its blockchain!

[DoM P]

I like how the NXT community is pushing for an attack on its blockchain!

The Nxt community knows the Nxt blockchain doesn't have much to do with APEX.
Let's have it...

[kiklo]

But killing zombies is a lot of fun.

so is this

c'mon, APEX's price hasn't moved, and they didn't even have checkpointing servers (not that it would have helped, but It shows they didn't care or they were not prepared to run a coin).
I like to think I'm not destroying work, I'm building a more secure future. For our kids! would anybody think about the children !?

Hey cynicSOB,
what do you think about that: https://bitcointalk.org/index.php?topic=897493.msg10208018#msg10208018

Would it lower your success?

Cheers,
Ray

I think it wouldn't work: how can you tell an attack from normal use? if you have 2 chains that look similar but have one conflicting tx you have to choose one, there's just no way to avoid that. How do you define which is the "good" one? an attack may not even change the difficulty

The theory behind it is , it blocks both Normal Use or Attack of anything that might have the potential to stake over the % where the attack could happen. So in theory if there are 2 blocks, anything over the % line is blocked, but anything below the % line is accepted normally, and if you have 2 block come in 1 over the % line , and 1 under , the one under it would be accepted , complete opposite of its normal behavior.
If both blocks are under the % line, then it stakes normally. If both are over the % line , then both are blocked for another block under the % line.
Kind of like we have to give the prize to the 2nd or 3rd place runner, cause the 1st one was too fast.
It would affect staking for a lot of users, but would be worth it, if it could remove the 51% vulnerability for all POS coins.

Thanks,
Kiklo

[cynicSOB]

Kind of like we have to give the prize to the 2nd or 3rd place runner, cause the 1st one was too fast.
It would affect staking for a lot of users, but would be worth it, if it could remove the 51% vulnerability for all POS coins.

but you didn't solve anything... how do you know the 2nd or 3rd place runners are not the attacker? maybe he placed the first one to trick you into selecting the second one which is also his...

The chance of forging a block does not depend on how long it is since the previous block you forged. It depends on what fraction of the forging coins you have.

It depends on both. Your "hit" is calculated by multiplying your stake by the seconds since the last block.

What is your criteria for a successful attack?

this one is easy: either successful double spend or proving you have 10 (or whatever is required for confirmation) blocks in a row.
I did both for Apex.

[kiklo]

Kind of like we have to give the prize to the 2nd or 3rd place runner, cause the 1st one was too fast.
It would affect staking for a lot of users, but would be worth it, if it could remove the 51% vulnerability for all POS coins.

but you didn't solve anything... how do you know the 2nd or 3rd place runners are not the attacker? maybe he placed the first one to trick you into selecting the second one which is also his...

Thanks for responding, it would not know if the 2nd or 3rd place or 15th are attacks or normal,
it would only block everything that could be a potential attack , which means some normal stakers would also have their stakes blocked , if it had the potential and was over the % line and only allow stakers below the % to ever add any stakes.  Basically can not a filter using some calculations involving total # of coins and current difficulty be created to block potential attacks , even though it will also block some normal stakers.
That is what I was wondering.

Thanks,
Kiklo

 

[Testing Crypto]

He is using testnet as well. In fact he's just been working on NXT testnet at the request of many. I'm guessing you haven't read the thread thus far. He's doing a good job so far and lots of people are interested in seeing more of his attempts.

This type of thing is critical for PoS to really be considered a robust consensus system. Actually it's not great that it's taken this long for some good public stress testing/attacks. Anyone interested in the idea of PoS should be supporting people who can help strengthen it.

I'm not saying that I fully agree with how this "stress testing" is being done, but am glad to actually see it done sooner than later. Just imagine if this wasn't pointed out, flaws of P.O.S. in 100's of coins till there was Millions of people holding 100's of wallets not opened for months? Not many people can sit back & think about how many things in the world are "stress tested", creating something more reliable & stable. Just think about a world without "stress testing", the wheel would still be just a wheel & break from the slightest bump

// Not that wording would change much, but maybe replace "attack < stress testing for the public".

[Daedelus]

Bolded

Some simple questions about nxt so I don't have to look them up:

1) Is transparent forging really implemented? I thought not

The part that allows you to predict the next block is, you should be able to confirm this yourself using the tool above. TF is several interlocking bits so you can't really pick a point and say "that's when TF began". And not all of it is implemented yet.

2) Is https://bitbucket.org/JeanLucPicard/nxt the latest source code? if not, where do I find it? do I really have to decompile Java?

I believe so. Every version ships with the source code in the 'src' folder so you could use that for the latest version. Not sure it is in as usable format though

3) If account A transfers some amount to account B (which was already verified) at block N, can B use that stake to forge block N+1? if not, when will it become part of B's effective balance?

You have to wait for N+1440 blocks before it counts in B's effective balance again. You can't broadcast a transaction, move the Nxt, broadcast again in the next block in a new account, move, broadcast move etc. using the same NXT.

4) Is re-leasing possible? if A leases to B, can B re-lease to C without approval from A?

No

5) which is the best place to ask these questions? nxtforums? which section?

Nxtforum is best.

Technical general questions: https://nxtforum.org/general/
Transparent Forging: https://nxtforum.org/transparent-forging/


Thanks!

You're welcome. Keep it up!

So, found any holes the Nxt devs missed in their implementation?  

[cynicSOB]

So, found any holes the Nxt devs missed in their implementation?  

I think so... but as I said, I need some time to implement the attack and completely understand the implications... attacking still requires a lot of stake, looking how to improve that...