delete

[doublec]

you can only go back to the last checkpoint. are there any in namecoin?

It looks like there are none.

[1714881158]

1714881158

[1714881158]

1714881158

[1714881158]

1714881158

[Lolcust]

Given how recent run-in on Geist proceeded (and assuming it was indeed EXpress, which he implied he was), I suspect he has more in mind than just crude double-spend.

Perhaps  someone should log/record the attack to increase its educational value ?

[jtimon]

In the name of science...

How is this in the name of science? What are you going to discover that you can't in a test chain?
In the name of science I ask you not to do it. If you don't do it we will very soon see:

1) A chain changing its rules for acceptable blocks. Bitcoin may need this in the future.
2) An existing chain changing to merged mining instead of starting with it.
3) Merged mining finally at work !!

I you're successful destroying namecoin, we will have to wait for another currency with merged mining from its beginning.
This is not in the name of science. You don't care to attack other people's good efforts. As DavinciJ15 says, you're a bad person.

Don't you like decentralized DNS?
Don't you like merged mining?
Is this all bullshit and you're just buying cheap nmc?
Seriously, what's wrong with you?

[Lolcust]

Technically,  as long as the attack did not start NMC devs could go and make a quick checkpoint (or several lol)

Fo instance, namecoin's block 18931 has hash of


getblockbycount 18931

Code:

    
    "hash" : "0000000000008adad642dd9567dcee1e59f7121881bfc6db66db1177bc251049",
    "version" : 1,
    "prev_block" : "000000000000a815eaf81de859f874b78c488c7dc2d0f138c3be2e506d819022",
    "mrkl_root" : "81e9d77d39ef85a046cbcff76c9cf0d23d6afb9d3ae11c5102c948c3693ad569",
    "time" : 1315842654,
    "bits" : 453030505,
    "nonce" : 817411455,

See? that was easy.

Now, someone could easily make a checkpoint "before it's too late"

[memvola]

if people were willing to accept some centralization until hashing power got to a "safe" level it could be solved by a central authority publishing block-chain checkpoints every X blocks.

How about doing it manually locally? How would it affect the result (assuming there's an attack) if majority of the nodes add arbitrary checkpoints (say, 120 blocks back)? Is it better if the checkpoint is the same on all the nodes but it's old, or heterogeneous but recent?

[gene]

There is no legitimate reason for threatening namecoin. Even children understand that just because you can do something, that doesn't mean you should.

An attack like this shows nothing that we don't already know. Why not try to learn something that we don't know? For instance, let's see if merged mining really works on production chains. Why not put the hashing power towards something constructive? Show everyone that you're not a vandal - that you're bigger than that; mine namecoins and and get it up to the merging mark.

Make an internet with distributed currency and free DNS a reality.

[twobits]

Technically,  as long as the attack did not start NMC devs could go and make a quick checkpoint (or several lol)

Fo instance, namecoin's block 18931 has hash of


getblockbycount 18931

Code:

    
    "hash" : "0000000000008adad642dd9567dcee1e59f7121881bfc6db66db1177bc251049",
    "version" : 1,
    "prev_block" : "000000000000a815eaf81de859f874b78c488c7dc2d0f138c3be2e506d819022",
    "mrkl_root" : "81e9d77d39ef85a046cbcff76c9cf0d23d6afb9d3ae11c5102c948c3693ad569",
    "time" : 1315842654,
    "bits" : 453030505,
    "nonce" : 817411455,

See? that was easy.

Now, someone could easily make a checkpoint "before it's too late"

Someone should get multicoin's namecoin config file updated with this and the new merged mining start block.

[Lolcust]

Given how recent run-in on Geist proceeded (and assuming it was indeed EXpress, which he implied he was), I suspect he has more in mind than just crude double-spend.

Perhaps  someone should log/record the attack to increase its educational value ?

So you noticed the 20,000 blocks in 73 minutes huh?

Yup.

But I am more concerned about some other aspects of your run-in, not the fact that you've earned yourself several hundred thousand geists (and it seems to me, scared away a certain exchange where you could have sold them ), so please elaborate on the methods you utilized (PM if you feel reluctant to discuss in public at this point)

[MaGNeT]

I'm mining Namecoins @ Coinotron.com now, just to protect Namecoin against the attack.
More people want to join?

And yes, I know it's mining at loss but every coin is at loss atm so I don't really care...

[nrd525]

I wonder if this attack would be legal?

[dishwara]

Good luck.
I hope this will help us to improve Bitcoin & other crypto currencies further.

[Maged]

Wait... I thought that Namecoin had all of the protection measures that Bitcoin had, and that these exploits were only possible due to the changes to the difficulty code the other chains made?

The great news is that, even if this does work, an emergency block-chain lockin can be made. That being said, it's literally the last line of defense.

[Lolcust]

Wait... I thought that Namecoin had all of the protection measures that Bitcoin had, and that these exploits were only possible due to the changes to the difficulty code the other chains made?

The great news is that, even if this does work, an emergency block-chain lockin can be made. That being said, it's literally the last line of defense.

Well, "other" chains are just "more" vulnerable (though IMVHO bitcoin's window of 2h is, all things considered, a mite too wide).

However, if I understand the 51 issue correctly, you can enforce arbitrary rules when you have 51+, so whether or not a given failsafe is present becomes a moot point.

[John Tobey]

Perhaps  someone should log/record the attack to increase its educational value ?

Abe is designed to record it, though it currently lacks a web interface to quickly look for chain splits.  The slow way is to look for blocks with more than one "next block" and follow the "next" links.  Anyway, I look forward to some nice test suite material.

If the 51% attack is successful, you won't have to worry about merged mining as it won't happen.

I'm scratching my head over this conclusion.  If you publish a block 19200, merged mining happens, and very likely the whole network overpowers the attackers.  So you must plan on indefinitely preventing the chain of greatest work from having over 19200 blocks.  Perhaps by messing with the difficulty, shoot it up high for a low-numbered range, so high that the new chain has greater work than the current one, so you push the current block number back a few thousand.  Then stop mining the tip and let the public network dither (as it's currently doing) while you prepare your next "longest chain" at an even lower block height.  Etc.

Still a lot of ongoing work and expense, though admittedly, I have to read up on the ArtForz exploit you mentioned.  And if the Namecoin developers come up with a fix (even something as simple as checkpoints or MM start block tweaks) you will declare victory and move on, right?

I don't think it's a bad attack.  Thanks for giving everybody the heads-up.

[322i0n]

Does anybody else think this anouncement is just a crude attempt to drive the price down.
announcing this does not increase but reduce the likely hood of a successful double spend.

It might encourage more NMC miners to commit more hash to the network thus creating coins quicker plus cause panic selling.

[Raoul Duke]

Does anybody else think this anouncement is just a crude attempt to drive the price down.
announcing this does not increase but reduce the likely hood of a successful double spend.

It might encourage more NMC miners to commit more hash to the network thus creating coins quicker plus cause panic selling.

Or to reduce Bitcoin hashing power...

Not me, but I do work for Apple in Cupertino LOL...

^^ Your assholeness is explained... Learnt it with your boss, hein?

[MaGNeT]

Does anybody else think this anouncement is just a crude attempt to drive the price down.
announcing this does not increase but reduce the likely hood of a successful double spend.

It might encourage more NMC miners to commit more hash to the network thus creating coins quicker plus cause panic selling.

It could be, but I'm not taking any chances.
And by the way, if we all mine them for a few days, merged mining will do the rest.

[doublec]

We notified the exchanges to prevent monetary loses by these guys. The purpose of this is knowledge and not to steal.

Notifying exchanges doesn't prevent monetary loss. If the rewrite occurs then the exchange will end up with a different balance of namecoins compared to the balance held in the exchange database. Loss will occur due to the exchange losing the coins but still having the users balance reflect what they originally deposited. Even adjusting for the blockchain reorganization the exchange loses in time and effort to fix up transactions and deal with irate users who have lost coins.

The only way for an exchange to avoid this is to close.

[Seraphim401]

I really hope this is just a stunt to get more people mining namecoins out of sympathy.
This way we get to 19200 quicker.If not, I hope the attacker miners catches fire!  
  

[doublec]

Using ArtForz modified exploit we are going to generate 5000 blocks in a matter of minutes, confirm them and commence experimenting.

Didn't artforz's exploit rely on the geitz method of calculating difficulty adjustments using the times on blocks? I don't think namecoin does this does it?