Preparations for August potential chain split

Support
1

As most of you will have followed there is some risk that with the first of August Bitcoin might get disrupted with a potential chain split.

See here a good summary about the risks and background:
https://bitcoin.org/en/alert/2017-07-12-potential-split

As Bitcoin is used for the security of the trade (MultiSig, security deposit) and of course for the transfer of one side of the trade we are exposed to the risks of such a situation.

It is not clear yet if we will halt trading (arbitrator revokes so nobody can create new offers or take offers) or if we only display a warning alert. We will decide when we get closer to the event and have better estimations for the risks that a chain split will happen. But we have to do that at least 1-2 weeks before as trades need to complete before 1.8.

Bisq used BitcoinJ and BitcoinJ is following automatically the longest PoW chain, so there is no easy way to force it to one specific chain. Even if there is it might be that your trading peer want to use the other chain and the Multisig deposit transaction will likely fail or the situation will get at least very messy.
Also if we would be able to force the whole Bisq network to one chain there is th open risk that transcactions don’t get confirmed for a very long time and in the worst case that the chain gets attacked by a reorg and past transaction would simply disappear (though the fiat or altcoin has been settled). So that is an in-acceptable risk and we cannot leave it to the users as some will simply be not sufficiently informed and might lose money.

We added with 0.5.0 the support for alternative currencies as base currency (Litecoin, Doge) and that will be the recommended way how to continue trading without being exposed to the risks of a chain split on Bitcoin.

We will add also the alternative Bitcoin currencies to the client so people can trade those on the other base currency markets (they are like altcoins out or system, transfer is on the users own risk).
The names and ticker symbols are not decided and we will probably follow that community suggests, in doubt if there is no clear distinction (there is expected a hard fight which one is the real Bitcoin) we will consider the unchanged Bitcoin as the original Bitcoin and any chain which have changed anything with a new name (e.g. UASFCoin, BitMainCoin, SegWit2xCoin,…).

Please be aware that trading and transferring any kind of Bitcoin in that time will come with a high risk and Bisq will not support you if you have issues in that area. It is completely out of system like any other altcoin and you are solely responsible to understand the technical implications and risks if you transfer one of the Bitcoin variants in such an event.

Hopefully miners consider to avoid the potential damage and activate SegWit before of 1.8. and then we will have very little risk that a fork will happen and has chances to succeed.

I personally am supporting both the unchanged Bitcoin (with BIP 149) and UASF (BIP 148).
I consider the block size debate as a vehicle for a political attack and attempt to obtail control over Bitcoin and reduce the level of decentralization and censorship resistance of Bitcoin.

I also consider the block chain congestion of the recent months as a result of a coordinated tx spam attack for increasing pressure on the political side as well as increasing the mining fees. As one can see in statistical data block chain congestion has not happened anymore since the SegWit2x parties have signed the NY agreement. Seems likely it was part of the deal to stop those attacks. That raises the question if those miners who are getting paid very well with mining rewards and fees for securing the network are working in the interest of Bitcoin and its uses or if they are motivated by other agendas.

I am aware that some of our users have other opinions and that is or course perfectly fine.
I just want to emphasize that Bisq will not act like other exchanges “neutral” or try to make opportunistically business with everybody but will take side and I see that aligned with our core values for decentralization, protection of privacy and censorship resistance.
If we accept politically motivated damages to those areas to Bitcoin we would also accept it to Bisq.
So for me it is a logical consequence to take a clear position instead of being opportunistic.

5 Likes
3

I beg to disagree. I deeply dislike how UASF tried to impose their view and then they complain the other proposal for reacting the same way they do.
IMHO the real danger is setting the precedent of people succesfully forcing consensus like the so-called-soft-fork gang.

4

Thanks a lot for sharing your view.
I agree with you: not to be opportunistic is the right choice.

1 Like
5

There is an interesting claim to be aware of, that the transaction spam was coming form a AlphaBay mixer, and stopped when they change to a new mixer. but I don’t know the details…
I wouldn’t put in past the spammers-that be to waist their funds on spamming in this instance, of course, and the timing of the ending of the congestion is suspicion, thought this attack was outstandingly expensive relative to the previous ones.

6

Do you have a link?

Tx spam is only expensive if it ends up in the blockchain. A miner cartel can agree to not put certain txs in the chain, as well with tiny fees and double spends you can prevent it on your own. Filling up mempool is free of cost. Thats weak point in the economic model of Bitcoin. It a distributed resource and hurting it does not cause costs.

7

Rising the price of tx confirmation is free, since Bitcoin protocol acts like an auction in that scenario and every auction has that problem. However increasing the confirmation time shouldn’t be, because either some other transactions need to fill the space, which costs miner fee for those transactions, or you need to mine blocks that aren’t full, which costs the miner fees you are missing out on.

Since we saw an increased confirmation time on lower fee transaction, there ware either more transactions that ware placed in the blocks (which would cost to make artificially) or blocks weren’t full (would also cost on missing out). However it could still be profitable in indirect ways.

If accepting the lower fee transactions in the same time would make rest of the users not increasing the fee (even tho users might see other higher fee transactions in the mempool) then the higher fee on other txs could be enough to cover the cost on completely missing out on other lower fee txs. It is important to mention that this scenario could still be unprofitable even if the attacker has the majority of the hashrate, but is not willing to orphan blocks of other miners, since the other miners could accept the lower fee blocks,even tho the confirmation time would increase in proportion to how much hashrate the attacker has (in this case you could see who is purposely not accepting lower fee txs and how much hashrate they probably have).

Other attack would be to make higher fee txs and fill the attacker mined blocks with them up to a certain point. An attacker could just fill them up to the amount of free space he had and not lose any money. However in practice this would not increase the confirmation time for lower fee txs. So as in the first attack scenario profitability depends on the fact that pushing some lower fee txs out of the blocks, the fee increases for the rest and this could be enough. The difference in these two attacks is just that in the second one blocks appear naturally full and does not affect the profitability of the attacks.

Of course none of the profitability matters if the gain is outside of the current network, or in other words, if these attacks influence the decisions on how the network scales and evolves or even if it maybe stops certain people or programs from sending txs. Either way, such an indirect gain from disruption of the network and it’s usability would classify such an (still just hypothetical) attack as an DOS attack. And such attacks could in general be for many possible reasons.

1 Like
8

Hi All!
Looks like no any split will happen.
At this moment only 19 blocks left until rejecting non SegWit blocks.
Bip 91 win, so everybody can relax…:slight_smile:

9

Yes hopefully.

There are still some open risks, see: https://www.reddit.com/r/Bitcoin/comments/6okd1n/bip91_lock_in_is_guaranteed_as_of_block_476768/dki17oc/

Here is a good page as well for monitoring chain splits:
https://www.btcforkmonitor.info/