Atomic swap? Objection: I don't think it means decentralized coin exchange.

[aliashraf]

We are hearing a lot about atomic swap (= cross chain transaction processing), specially in the past few weeks.

From a theoretical point of view, I can confirm it is a good start for solving the most important problem in the crypto world nowadays: the need for a decentralized exchange, but I doubt it can be considered a solution or we are close enough.

My biggest concern is scalability, obviously, but there is other challenges as well. I was just wondering if anybody has been studying the subject before and can share her/his ideas with me?

Actually I have been through a very bad experience with a centralized exchange (Bittrex) lately and it has escalated my old anti-centralization sensitivities and allergies about exchanges  .

So, I'm here now, determined to participate more actively (by means of coding, investing, whatever) in shutting the doors of these scammy businesses (bittrex, poloniex, ..., f*ex ) forever and I have a couple of lines of codes,  some coins and a few ideas to share.  

[1714940952]

1714940952

[1714940952]

1714940952

[1714940952]

1714940952

[687_2]

My biggest concern is scalability, obviously, but there is other challenges as well.

For XCATs you'll need to run a full node for each cryptocurrency you are trading, e.g. you'll need to run both Zcash and Bitcoin core if you want to exchange those two cryptocurrencies. You'll also need to run some middleware that connects everything, see https://github.com/zcash-hackworks/zbxcat

Merchants that want to accept multiple cryptocurrencies and convert them into Bitcoin for e.g. paying taxes in Bitcoin will probably have fairly significant disk storage requirements.

[Taras]

My biggest concern is scalability, obviously, but there is other challenges as well. I was just wondering if anybody has been studying the subject before and can share her/his ideas with me?

Have you heard of the lightning network? My understanding is that atomic swaps would use that. So, it's highly scalable. You can buy litecoin 100 times over the span of a month using one bitcoin transaction and one litecoin transaction.

[aliashraf]

My biggest concern is scalability, obviously, but there is other challenges as well. I was just wondering if anybody has been studying the subject before and can share her/his ideas with me?

Have you heard of the lightning network? My understanding is that atomic swaps would use that. So, it's highly scalable. You can buy litecoin 100 times over the span of a month using one bitcoin transaction and one litecoin transaction.

The way I have understood Lightning, it is focused on micropayments in its wierd approach: channels that work like buffers and are 'flushed' periodically while being managed by specialized nodes , ... (believe it? how some people are so excited about this proposal?) I don't think it can help cryptocurrency at all,  specially it has nothing ,to do with chain to chain transactions, unless the two chains may be basically very similar (like ltc and btc) and they both implement lightning, and lightning has been developed for this use case, .... long, long time to go. I'm not so patient

Actually I'm talking about something to be done in a couple of months and even faster. It is agile era ... lightning is out of gas.

[haltingprobability]

The way I have understood Lightning, it is focused on micropayments in its wierd approach: channels that work like buffers and are 'flushed' periodically

You've simplified to the point of over-simplification. The "buffer" (payment channel) has an expiration that is pre-set. The payment channel can be utilized to make many small payments until it expires. It's not really a periodic buffer like a hardware buffer or a software queue, because the channel simply closes for good once it expires.

while being managed by specialized nodes , ... (believe it? how some people are so excited about this proposal?)

This is a skewed perception. Payment channels do not need anybody to "manage" them at all. LN adds a way to "chain" payment channels together but nobody in a LN is "managing" anything. One way you can utilize LN is to build a hub-and-spoke (or "star topology") payment model that allows many individual users (spokes) to have an open payment channel with one hub so that they can all make payments to one another without having to open a zillion payment channels on the blockchain. "So, we're back to centralized payment systems." Not quite. Since any individual can open a payment channel with any other individual, the central-hubs of hub-and-spoke style LN networks cannot "monopolize and then jack up fees". They will have no choice but to charge going rates, otherwise, users will just walk away and do it manually.

I don't think it can help cryptocurrency at all,

On the contrary, cryptocurrency will never receive widespread adoption without it.

[Wind_FURY]

My biggest concern is scalability, obviously, but there is other challenges as well. I was just wondering if anybody has been studying the subject before and can share her/his ideas with me?

Have you heard of the lightning network? My understanding is that atomic swaps would use that. So, it's highly scalable. You can buy litecoin 100 times over the span of a month using one bitcoin transaction and one litecoin transaction.

In theory, yes. But we if do not see that it is possible in practice, we have nothing. I would rather see Bitsquare developed further and used more. It is a more practical solution in the state of the technology that we have now in my opinion.

[d5000]

I am also very interested in the topic and eager to get informations about actual "real" tests that are going on. According to this Spanish article there were tests with Bitcoin, Zcash, Litecoin, Decred, ARK and Vertcoin, some weeks ago.

The traditional atomic swap model supposes one multisig transaction per chain. That obviously isn't enough for high frequency trading but would enable the typical use case of a trader that wants to "ride" bullish and bearish cycles of distinct coins. For higher frequencies, one could use LN, or maybe even a specialized pegged sidechain in the future (RSK?). For LN, however, I think it would be used mainly for small exchanges of less than 50-100 USD because of the slightly increased risk.

By the way: There is a atomic swap system that works for two years now used only by two small altcoins but supposedly could be adapted to Bitcoin-based coins. I have recently downloaded both blockchains to test it soon, if I am successful I'll report results here.

For XCATs you'll need to run a full node for each cryptocurrency you are trading, e.g. you'll need to run both Zcash and Bitcoin core if you want to exchange those two cryptocurrencies. You'll also need to run some middleware that connects everything, see https://github.com/zcash-hackworks/zbxcat

Thanks for pointing me to that software, will try to test it. A question arises: Would it be possible to use SPV wallets like Electrum for atomic swaps? If yes, then the storage problem would be solved ... Another idea for mass adoption could be if "hybrid semi-centralized" wallets like blockchain.info included an atomic swap feature, as they could manage "chain handling" for end users.

Merchants that want to accept multiple cryptocurrencies and convert them into Bitcoin for e.g. paying taxes in Bitcoin will probably have fairly significant disk storage requirements.

If SPV wallets are not possible, Bitcoin 0.11+ style pruning could be the solution for that - a merchant then could easily accept up to about 100 currencies with ~1GB storage use for every chain.

[coinmachina]

My biggest concern is scalability, obviously, but there is other challenges as well. I was just wondering if anybody has been studying the subject before and can share her/his ideas with me?

Yeah compared to centralized exchanges, cross chain swaps would not scale very well, assuming you already have the your money at the exchange.

If you don't normally store your money on an exchange and first would have to send your coins to the exchange, exchange it, and then withdraw your new money from the exchange, an atomic cross-chain swap would probably be more convenient. And of course more secure.

[aliashraf]

The way I have understood Lightning, it is focused on micropayments in its wierd approach: channels that work like buffers and are 'flushed' periodically

You've simplified to the point of over-simplification. The "buffer" (payment channel) has an expiration that is pre-set. The payment channel can be utilized to make many small payments until it expires. It's not really a periodic buffer like a hardware buffer or a software queue, because the channel simply closes for good once it expires.

not bad for a fresh start in a forum, but on the contrary you are complicating a trivial concept to the point of over-complication IMO.
You are just suggesting it (a payment channel in LN terminology) is not a 'buffer' because it has an expiration date, right? But I think buffers can have a auto-flush feature triggered by a clock (block height in a chain, for instance) and they are still nothing more than a buffer.
I understand coupling a decentralized ledger with such a data structure is a very hard work and needs a lot of tricks and protocol developments but in the end we just have a simple data structure in hand with a very limited set of practical use cases, I'll come to this in my next paragraphs.

while being managed by specialized nodes , ... (believe it? how some people are so excited about this proposal?)

This is a skewed perception. Payment channels do not need anybody to "manage" them at all. LN adds a way to "chain" payment channels together but nobody in a LN is "managing" anything. One way you can utilize LN is to build a hub-and-spoke (or "star topology") payment model that allows many individual users (spokes) to have an open payment channel with one hub so that they can all make payments to one another without having to open a zillion payment channels on the blockchain. "So, we're back to centralized payment systems." Not quite. Since any individual can open a payment channel with any other individual, the central-hubs of hub-and-spoke style LN networks cannot "monopolize and then jack up fees". They will have no choice but to charge going rates, otherwise, users will just walk away and do it manually.

Oh you totally misunderstood my criticism, I have not issued any centralization accusation, not at all! I'm talking about specialized nodes (hubs) which are needed for an ad-hock payment to take place between an arbitrary couple of wallets. And this hub is managing the channel for sure, even if we don't name it this way.

I don't think it can help cryptocurrency at all,

On the contrary, cryptocurrency will never receive widespread adoption without it.

And here is your biggest mistake and the very important point you have overlooked: cryptocurrency adoption is nothing to do with buffers (auto-flushed or whatever) because the main use case is not recurring payments in the real world, and it is just what LN is solely about.

I prefer to discuss LN in another topic, obviously LN is not about inter-chain atomic swaps and we can remain focused on the main subject, which is whether atomic swap can improve state of the art in decentralized exchange development or it can not.

[J. Cooper]

I agree with you that centralized exchanges have to go. Even though I never have had any problems with them, I still feel that they're undermining the essence of decentralization. How can a coin be truly decentralized when it can only be traded (at a large scale) one an exchange that is owned by a corporation.
I was very pleased when I read up in the concept of atomic swaps but seemingly you don't. What are the downsides of atomic swaps that we won't be able to fix in the future?

[arisen725]

Atomic swaps are definitely a powerful tool that can be used to implement decentralized exchange. Personally I hope we will see more efforts to connect ledger systems in this way, and that it will make cross blockchain trading normal sooner than later.

The technology itself is somewhat complicated, especially between blockchains that use entirely different cryptographic schemes. So give it time to mature, and eventually I think we will be trading outside the central exchanges more than on them bringing less manipulation of the order books, and more resilience.

[687_2]

Thanks for pointing me to that software, will try to test it. A question arises: Would it be possible to use SPV wallets like Electrum for atomic swaps? If yes, then the storage problem would be solved.

I think not, because you are dealing with a complete stranger- perhaps an adversary. So you must validate for yourself that you have e.g. 1 confirmation and you really can't trust any external party for that information. There is also the question of adversarial forks, so this would be the only way to guarantee you are running on the chain you think you are, and that your trading partner is not trying to use a slightly different chain. But I'm sure these services will be made available anyway, and people will use them without thinking about it  

If SPV wallets are not possible, Bitcoin 0.11+ style pruning could be the solution for that - a merchant then could easily accept up to about 100 currencies with ~1GB storage use for every chain.

You'd probably only be able to trade with other people who have the same pruning settings on their chains, but this is just a wild guess. The properties of the trading pairs are a key consideration, and would have an impact on your storage requirements. For example, with Monero you don't need to download the blockchain, you can connect to a remote node. So additional storage would not be required if you're swapping XBT/XMR. But a ETH/XBT swap is a totally different story, you'd have to have massive storage for that swap.

I think by their very nature, XCATs are going to be much slower than the centralized services that are available today. I think each swap, at least where Bitcoin is involved, will take at least 10 minutes. I don't really imagine XCATs will be used by "traders/speculators", but rather by people who genuinely wish to exchange 10 XBT for 500 XMR, wishing for complete privacy, near zero cost and assurance that the trade will not fail.

[Ucy]

We all have this problem of abandoning great ideas once we have temporal relief to bad problems. If we had stuck with developing better alternatives to Centralized exchanges we would have come up with great decentralized exchanges by now.
It is good to be consistent in all we believe in until we have good reason not to.

I have gone through few articles on Atomic Swap and found them very interesting.

[Snub]

Yeah, there is always hard to know for sure if it is a good idea to make something like Atomic Swap, but as for me - as more cryptocurrencies we have - as more we need something like it.
So, we have to give it a chance, like we did it to bitcoin itself many years ago)

[aliashraf]

I agree with you that centralized exchanges have to go. Even though I never have had any problems with them, I still feel that they're undermining the essence of decentralization. How can a coin be truly decentralized when it can only be traded (at a large scale) one an exchange that is owned by a corporation.

Lucky you, not having problems with centralized exchanges, personally, I have been scammed by Bittrex a couple of weeks ago and just like thousands of other people, in the excuse of KYC or something (never been officially declared).

I was very pleased when I read up in the concept of atomic swaps but seemingly you don't. What are the downsides of atomic swaps that we won't be able to fix in the future?

Sure, atomic swap technology is gradually succeeding and we have news in this field almost on a daily basis. It is great, we need it, and there is no downside here to have a protocol to do some fascinating distributed ledger transaction processing, but, unlike some proponents of this technology, I'm not sure it can solve the decentralized exchange problem by its own even once it is available, full fledged and open source.

The main challenge (as I have mentioned before) is scalability, and it is a snow ball situation:

Atomic swap, if going to be used directly as an infrastructure to facilitate peer-to-peer coin exchange in the very first steps imposes large storage and hardware requirements to both peers. As @d5000 and @687-2 have discussed about in previous posts here. Now it is just for beginning. Suppose in some point we have overcome this challenge and got to run our nodes, supporting AS. Then what? It will escalate transactions and we will have more scalability problem to overcome and so on. So, obviously we have a limit here, solving the problem raises the scale of the problem and no one will be satisfied ever.

Another major concern is that Atomic swap is just an infrastructure, it can not be projected directly to the problem domain. Trade is a complicated socioeconomic behavior, one can not address it with a device. In the exchange case you need a marketplace, an order book, a matching rule, ...

[Wind_FURY]

I am also very interested in the topic and eager to get informations about actual "real" tests that are going on. According to this Spanish article there were tests with Bitcoin, Zcash, Litecoin, Decred, ARK and Vertcoin, some weeks ago.

The traditional atomic swap model supposes one multisig transaction per chain. That obviously isn't enough for high frequency trading but would enable the typical use case of a trader that wants to "ride" bullish and bearish cycles of distinct coins. For higher frequencies, one could use LN, or maybe even a specialized pegged sidechain in the future (RSK?). For LN, however, I think it would be used mainly for small exchanges of less than 50-100 USD because of the slightly increased risk.

By the way: There is a atomic swap system that works for two years now used only by two small altcoins but supposedly could be adapted to Bitcoin-based coins. I have recently downloaded both blockchains to test it soon, if I am successful I'll report results here.

I did not know that atomic transfers were already invented two years ago. Thanks for sharing.

Which coins were used to do the transfers and why is the technology not widely used today? This is a breakthrough and no one is giving it any attention or giving the developer any recognition.

[d5000]

I think not, because you are dealing with a complete stranger- perhaps an adversary. So you must validate for yourself that you have e.g. 1 confirmation and you really can't trust any external party for that information. There is also the question of adversarial forks, so this would be the only way to guarantee you are running on the chain you think you are, and that your trading partner is not trying to use a slightly different chain. But I'm sure these services will be made available anyway, and people will use them without thinking about it  

So I interpret that it would be "possible" but potentially insecure. Well, for small amounts of less than (let's say) ~$100 (I think those make up the majority of altcoin trades) it would be a relatively "doable" solution if a swap-capable client checks various block explorers of different origin if the confirmations have happened.

You'd probably only be able to trade with other people who have the same pruning settings on their chains, but this is just a wild guess.

Interesting, but I "don't get it" for now.  Could you clarify why you think the pruning settings should be similar or give me a hint where I could research?

I did not know that atomic transfers were already invented two years ago. Thanks for sharing.

Which coins were used to do the transfers and why is the technology not widely used today? This is a breakthrough and no one is giving it any attention or giving the developer any recognition.

The coins that already can be used for atomic swaps are Burst and Qora. At least the Qora client has even a GUI option for that (see this article).

The reason I think it is not widely used is that both coins share a relatively similar codebase (they are Java coins inspired by NXT and use a smart contract platform called "Automated Transactions" for the atomic swap implementation) that is pretty different from the standard "Bitcoin clones" and Bitcoin/Script coins need an alternative implementation. Burst and Qora never were among the "top altcoins".

I think also the creator CIYAM deserves more credits for his breakthrough implementation. His solution is based on a concept invented by Sergio Demian Lerner and improved by Tier Nolan but it's the only implementation I know that is usable by non-technical users.

[687_2]

I think not, because you are dealing with a complete stranger- perhaps an adversary. So you must validate for yourself that you have e.g. 1 confirmation and you really can't trust any external party for that information. There is also the question of adversarial forks, so this would be the only way to guarantee you are running on the chain you think you are, and that your trading partner is not trying to use a slightly different chain. But I'm sure these services will be made available anyway, and people will use them without thinking about it  

So I interpret that it would be "possible" but potentially insecure. Well, for small amounts of less than (let's say) ~$100 (I think those make up the majority of altcoin trades) it would be a relatively "doable" solution if a swap-capable client checks various block explorers of different origin if the confirmations have happened.

Yes I think just like some merchants accept zero confirmations for coffee purchases, people can create XCAT systems that allow the user to take greater risks. But in XCAT systems your trading partner is likely an adversary..the relationship is the opposite of doing business with Bitcoiners in person, where we are excited to meet and do business with each other.

You'd probably only be able to trade with other people who have the same pruning settings on their chains, but this is just a wild guess.

Interesting, but I "don't get it" for now.  Could you clarify why you think the pruning settings should be similar or give me a hint where I could research?

If you prune to 550 and your trading partner prunes to 950, then she has a 400 block informational advantage over you. If she is able to introduce a serious network disruption (perhaps a hardfork attempt involving a lot of hashpower!?) during the trade she might be able to take your coins while giving nothing in return.

I would personally want to make sure that I'm not risking anything if some kind of serious instability arises, especially when trading larger amounts. So in many cases that means no pruning at all.

[DooMAD]

You'd probably only be able to trade with other people who have the same pruning settings on their chains, but this is just a wild guess.

Interesting, but I "don't get it" for now.  Could you clarify why you think the pruning settings should be similar or give me a hint where I could research?

If you prune to 550 and your trading partner prunes to 950, then she has a 400 block informational advantage over you. If she is able to introduce a serious network disruption (perhaps a hardfork attempt involving a lot of hashpower!?) during the trade she might be able to take your coins while giving nothing in return.

I would personally want to make sure that I'm not risking anything if some kind of serious instability arises, especially when trading larger amounts. So in many cases that means no pruning at all.

If you and your trading partner were the only two participants on the chain you were transacting upon, then it might be an issue, but if every other participant on that network has a copy of the blockchain, the person attempting to introduce the disruption or fork would only succeed in forking themselves off the network, so that attack vector wouldn't work.  It would take a significant proportion of network participants to be colluding in the act (basically what tends to be referred to as a 51% attack) to mess with your transaction in such a manner.  Otherwise no one could rely on SPV, where you don't keep a copy of the blockchain at all.

[n41r0j]

Atomic swaps have happened for quite a while actually. I am part of the testing team at Komodo Platform, where we have been actively swapping around 80 coins for a couple of months now. They have built a decentralized exchange, including decentralized orderbooks, decentralized order matching, etc. An easy to use GUI is in BETA now.

Some articles that probably explain it better than me:
https://blog.komodoplatform.com/everything-you-need-to-know-about-atomic-swaps-and-how-komodo-is-advancing-the-technology-cadaec50da7c
The topic where the current swap protocol is explained and 'invented': https://bitcointalk.org/index.php?topic=1340621.0
And the whitepaper covering all the details about the decentralized exchange BarterDEX: https://github.com/SuperNETorg/komodo/wiki/barterDEX-Whitepaper-v2

Thanks for pointing me to that software, will try to test it. A question arises: Would it be possible to use SPV wallets like Electrum for atomic swaps? If yes, then the storage problem would be solved ... Another idea for mass adoption could be if "hybrid semi-centralized" wallets like blockchain.info included an atomic swap feature, as they could manage "chain handling" for end users.

Merchants that want to accept multiple cryptocurrencies and convert them into Bitcoin for e.g. paying taxes in Bitcoin will probably have fairly significant disk storage requirements.

If SPV wallets are not possible, Bitcoin 0.11+ style pruning could be the solution for that - a merchant then could easily accept up to about 100 currencies with ~1GB storage use for every chain.

Swaps between two Electrum nodes have been successfully done for a couple of weeks now, too.