Page 3 of 7

Re: Namecoin is Prime for a 51% attack

Posted: Mon Sep 12, 2011 7:36 pm
by smoothie
I agree with Davinci on this and disagree with BitcoinExpress on this.

Dude if you want to test something do it on a test chain that DOESN'T involve people's investments.

Maybe merged mining should be implemented ASAP? That would crush this guy's idea in a minute.

Why is it taking so damn long anyway? Changes to other networks have taken merely days and at most a week. So why would merged mining take 2.5 months?

The fact that it took this long is why we are in this predicament with someone posing a threat to the overall network.

Re: Namecoin is Prime for a 51% attack

Posted: Mon Sep 12, 2011 7:44 pm
by johntobey253
I think starting merged mining now might solve more problems than it creates. But I do expect further attacks, and I doubt whether the merged mining code is exploit-free. For anyone who wants to start, I've created a version in GitHub, TOTALLY UNTESTED, that starts merged mining immediately: https://github.com/jtobey/namecoin/tree ... .61+mm-now Edit: removed and deprecated in favor of official VinceD release.

I will see about adding this to the versions tracked by Abe so we can watch the action. No promises, though. Update: done. Abe's behavior after a split is untested. :)

Re: Namecoin is Prime for a 51% attack

Posted: Mon Sep 12, 2011 7:45 pm
by Davincij15
Seraphim401 wrote:What gives you the right to this to namecoin?

He has every right. It sucks but we need to deal with it not bitch at him for doing it. Just be glad he announced it instead of doing it secretively.

Re: Namecoin is Prime for a 51% attack

Posted: Tue Sep 13, 2011 1:14 am
by doublec
There was some discussion with ArtForZ in #namecoin recently about the threatened attack. It looks like the attackers plan is something like:

1) Use a variant of this flaw in bitcoin/namecoin to enable fast generation of blocks from block 10,000 up to the merged mining point.
2) Use their large pool hash power to mine from the merged mining point on to continue the >51% attack. They can do this 'for free' since they're earning bitcoins due to being past the merged mining block.

So some of the hash power they're using to overtake the chain is being done during the merged mining period where they are earning bitcoins making it essentially free for them to attack the chain. What could be done to help prevent it:

3. Add a checkpoint to the namecoin client immediately so they can't generate from 10,000-currentblock using the exploit from (1) above.
4. Now we're just dealing with a usual 'bad actor with >51%' issue. We'll be at the merged mining point so can start getting larger pools (Eligius?) to join in and help stop reading 51%.

ArtForZ suggesting fixing the flaw in (1) if we're going to be distributing a new checkpointed client in (3). Vinced, thoughts?

Re: Namecoin is Prime for a 51% attack

Posted: Tue Sep 13, 2011 2:10 am
by smoothie
All I can say is I have 4 GH/s I can use to help if need be. But if they (he) does succeed they will have destroyed the project altogether even if they only rewrote 9000 blocks.

I will have lost all the namecoins I've been mining and purchased.

I will also have lost the domains I have registered through the network.

I will be left with no choice but to move on from namecoin empty-handed and with a bad taste in my mouth and let others coming into the market to steer clear of the project (if he/they are successful).

We'll see... :?

Re: Namecoin is Prime for a 51% attack

Posted: Tue Sep 13, 2011 3:23 am
by paraipanakos
smoothie wrote:All I can say is I have 4 GH/s I can use to help if need be. But if they (he) does succeed they will have destroyed the project altogether even if they only rewrote 9000 blocks.

I will have lost all the namecoins I've been mining and purchased.

I will also have lost the domains I have registered through the network.

I will be left with no choice but to move on from namecoin empty-handed and with a bad taste in my mouth and let others coming into the market to steer clear of the project (if he/they are successful).

We'll see... :?
heh, just had a thought like this first time i saw this but hey i'm in for the ride, kudos to them for making it public

ArtForz himself gave us the solution too, fixings the blocks time-stamp from devel view and the second one that was not stated can be that every peer maintains it's already downloaded "ledger" and stops transacting.
Remember we know very little when it comes to bitcoin protocol but they will be majority for a small time-span and working on a fork after all. My client will definitively not be accepting a ~9000 block reset if i don't re-download the whole blockchain.
Having block hashes included into main bitcoin client was a good idea for bitcoin after all, heh
Hope vinced would look into it and don't freakout ppl, as long as one peer with the honest ledger (blockchain) exist on the network you will not lose your coins or domains.

Re: Namecoin is Prime for a 51% attack

Posted: Tue Sep 13, 2011 4:54 am
by Davincij15
smoothie wrote:All I can say is I have 4 GH/s I can use to help if need be. But if they (he) does succeed they will have destroyed the project altogether even if they only rewrote 9000 blocks.

I will have lost all the namecoins I've been mining and purchased.

I will also have lost the domains I have registered through the network.

I will be left with no choice but to move on from namecoin empty-handed and with a bad taste in my mouth and let others coming into the market to steer clear of the project (if he/they are successful).

We'll see... :?
I'm offering just under 0.00003000 BTC to save the network! Let people know. It's up and running right now just click on PPS and refreash the web page you will see your value of shares change to BTC.

You can still mine nmc but I keep them if you enable PPS.

GET out there and sell it.

People get paid PPS when a new NMC block is found.

We need to fight off this attacker to save the time and effort we invested in this.

Re: Namecoin is Prime for a 51% attack

Posted: Tue Sep 13, 2011 7:07 am
by tommy
Can we all agree that we will manually revert back to block 19100 as the last valid block (before the scheduled attack that starts on block 19101) so that we can preserve all the honest transactions that has already been made? I bought namecoins with real money and would hate to see them seize to exist if and when the attackers manage to revert the blockchain to approximately block 10,000, as they are claiming will happen.

My namecoins at this address would simply seize to exist: http://explorer.dot-bit.org/a/NEuKQ4uJP ... ieeFqLzYrd

There is real life value that is being lost, if we let this revert happen and continue from the forked blockchain. If the community lets this value be unaccounted for and therefore lost, I will consider it too risky to buy namecoins again until it has reached mainstream usage.

Can this be hardcoded into the namecoin client software? Please provide an updated client version that will revert back to block 19100 once this attack is over.

Re: Namecoin is Prime for a 51% attack

Posted: Tue Sep 13, 2011 7:12 am
by doublec
tommy wrote: Can this be hardcoded into the namecoin client software? Please provide an updated client version that will revert back to block 19100 once this attack is over.
Run a client that has a recent checkpoint and it'll refuse to accept any chain that was rewritten before that checkpoint. Are you able to do your own namecoin builds?

Re: Namecoin is Prime for a 51% attack

Posted: Tue Sep 13, 2011 7:13 am
by phelix
thirdlight wrote:So, 160 blocks to go. That's about 3 weeks.

Would you wait that long?
they could speed it up quite a bit with their hashing power. Would be weired to then throw away their own work, though.

does it help namecoin to run checkpointed clients without mining? are there any windows binaries for a namecoin client with checkpoints yet?

+1 to 19100 (lol, but I mean it)

are there any sites that display the namecoin hashrate yet (besides abe)? maybe I will put something up on bitcoinx.com