Bitcoin Forum

Bitcoin => Development & Technical Discussion => Topic started by: d5000 on February 19, 2018, 08:49:51 PM



Title: Sidechains - state of the art?
Post by: d5000 on February 19, 2018, 08:49:51 PM
Some time passed since I heard really something new about sidechains as a solution to the scaling problem. The technology seems to be "stuck". This article from January about NiPoPoW (https://www.coindesk.com/sidechains-breakthrough-almost-everyone-bitcoin-missed/) is one of the few news about the topic I saw lately.

I'm curious what the current state of the art is. Are there people here that know where I could search and look at, to know more about the current sidechain discussion? Are there active projects, e.g. working on testnet?

I know that Rootstock and Blockstream have implemented federated sidechains, but they're too centralized to be really called "solutions", for me. What I'm more interested in are (almost) trustless designs like Paul Sztorc's Drivechain (http://drivechain.info).


Title: Re: Sidechains - state of the art?
Post by: achow101 on February 19, 2018, 09:11:39 PM
The sidechains that Blockstream and Rootstock use are based off of the Elements Project (https://www.elementsproject.org/) (which was created by Blockstream). There is an test/demo testnet sidechain called Elements Alpha (https://www.elementsproject.org/sidechains/alpha/). I believe it is possible to use the Elements Project's software in order to create your own sidechain. I think these can be both federated and unfederated sidechains too, although I don't think any currently existing sidechains are unfederated.


Title: Re: (Alpha Elements) Sidechains - state of the art?
Post by: pebwindkraft on March 13, 2018, 06:21:44 AM
...
There is an test/demo testnet sidechain called Elements Alpha (https://www.elementsproject.org/sidechains/alpha/). I believe it is possible to use the Elements Project's software in order to create your own sidechain.
Just wondering if this project is under rework? I tried to get involved via the link on the webpage https://www.elementsproject.org/posts/announcing-elements/ (https://www.elementsproject.org/posts/announcing-elements/), but the link to the slack channel seems dead: https://chat.elementsproject.org (https://chat.elementsproject.org). (At least my Safari is showing as "could not connect").
Also the Freenode IRC: #sidechains-dev is very quiet, no info since last November in the web archives.

My problem is currently, that my node doesn't connect to any alpha server. After having compiled everything, and made the daemons working, it looks like the alphad blocks are not downloaded (in ~/.bitcoin/alphatestnet3). There is only one block, and the logfile reports: "alphad cannot connect to <servername>, unsupported network". Looks like there is some TOR required, cause the <servername> ends  with ".onion", but the installation manuals are quiet on this...

Maybe someone has an idea or a reference? (Also checked bitcoin.SE, to no avail)


Title: Re: (Alpha Elements) Sidechains - state of the art?
Post by: achow101 on March 13, 2018, 02:48:10 PM
Just wondering if this project is under rework? I tried to get involved via the link on the webpage https://www.elementsproject.org/posts/announcing-elements/ (https://www.elementsproject.org/posts/announcing-elements/), but the link to the slack channel seems dead: https://chat.elementsproject.org (https://chat.elementsproject.org). (At least my Safari is showing as "could not connect").
Also the Freenode IRC: #sidechains-dev is very quiet, no info since last November in the web archives.
AFAIK not many people are doing stuff with sidechains so it's fairly quiet. Elements is still being worked on, I don't know what the current state of it is.


Title: Re: Sidechains - state of the art?
Post by: d5000 on March 17, 2018, 05:13:49 AM
Thanks to your answers (sorry for the late answer, I forgot about that thread). I admit I expected a bit more, but the technology seems to move on very slowly. I just looked at the Github repo (https://github.com/ElementsProject/elements) of the Elements project and there seems to be some activity these days. However, I think they are related more to the "Confidential Assets" feature which seems to be unrelated to sidechains.

I believe it is possible to use the Elements Project's software in order to create your own sidechain. I think these can be both federated and unfederated sidechains too, although I don't think any currently existing sidechains are unfederated.
"Unfederated" sidechains are the technology I am most curious about. I have just visited this description site (https://www.elementsproject.org/elements/deterministic-pegs/) and at least in the described version of Elements a "Federation" seem to be the "standard" way to create the "deterministic peg".

But if someone has information about the Elements project being used for an "unfederated" sidechain I would be glad if some information about it is shared in this thread. It doesn't seem impossible, given the large number of Elements forks (https://github.com/ElementsProject/elements/network/members).




Title: Re: Sidechains - state of the art?
Post by: nicosey on March 17, 2018, 08:44:21 PM
Thanks to your answers (sorry for the late answer, I forgot about that thread). I admit I expected a bit more, but the technology seems to move on very slowly. I just looked at the Github repo (https://github.com/ElementsProject/elements) of the Elements project and there seems to be some activity these days. However, I think they are related more to the "Confidential Assets" feature which seems to be unrelated to sidechains.

I believe it is possible to use the Elements Project's software in order to create your own sidechain. I think these can be both federated and unfederated sidechains too, although I don't think any currently existing sidechains are unfederated.
"Unfederated" sidechains are the technology I am most curious about. I have just visited this description site (https://www.elementsproject.org/elements/deterministic-pegs/) and at least in the described version of Elements a "Federation" seem to be the "standard" way to create the "deterministic peg".

But if someone has information about the Elements project being used for an "unfederated" sidechain I would be glad if some information about it is shared in this thread. It doesn't seem impossible, given the large number of Elements forks (https://github.com/ElementsProject/elements/network/members).




"Confidential Assets" is part of elements alpha so are part of sidechains.


Title: Re: Sidechains - state of the art?
Post by: gexaha on April 18, 2018, 08:30:17 PM
New video from IOHK about sidechains:

IOHK Research | Dionysis Zindros, Sidechains
https://www.youtube.com/watch?v=Y5QUGqFQnWg


Title: Re: Sidechains - state of the art?
Post by: d5000 on April 20, 2018, 02:45:45 AM
New video from IOHK about sidechains:

IOHK Research | Dionysis Zindros, Sidechains
https://www.youtube.com/watch?v=Y5QUGqFQnWg
Thanks! Good introductory video. While I already knew most of the things Zindros spoke about in the first part, I enjoyed the part about Proofs of Proofs of Stake.

It's a bit a pity that he didn't profundize the part about NiPoPoW, although I then found this article (https://bitcoinmagazine.com/articles/sidechains-why-these-researchers-think-they-solved-key-piece-puzzle/) with a little explanation what makes NiPoPoW different from (interactive) PoPoW. In short (if I understood it right): the "old" PoPoW couldn't guarantee that a second communication round between sidechain and settlement chain wouldn't be needed, while NiPoPoW claims to guarantee that the proof can be included in a simple string, so the communication would consist of a single "round", what makes it more factible for sidechains. I have to investigate further, but the fog is beginning to clear for me ...



Title: Re: Sidechains - state of the art?
Post by: d5000 on October 13, 2018, 03:26:05 PM
It seems that the first test version of Paul Sztorc's Drivechain has been released at the end of September, without much news buzz:

http://www.drivechain.info/blog/first-release/

Usage tour: http://www.drivechain.info/blog/usage-tour/

I still have to investigate how they solved the problem of the missing Bitcoin Core functionality, but I consider this big news, even if it seems to be something like a testnet. Working drivechains would be the perfect complement for LN and the main chain.


Title: Re: Sidechains - state of the art?
Post by: alko89 on October 14, 2018, 08:27:02 AM
One thing about "sidechains" is that they don't really need to be blockchains. You could use a normal database and store hashes of that data on the mainchain. I think that is very efficient.

OriginTrail for example uses that approach.


Title: Re: Sidechains - state of the art?
Post by: Wind_FURY on October 16, 2018, 06:55:48 AM
It seems that the first test version of Paul Sztorc's Drivechain has been released at the end of September, without much news buzz:

http://www.drivechain.info/blog/first-release/

Usage tour: http://www.drivechain.info/blog/usage-tour/

I still have to investigate how they solved the problem of the missing Bitcoin Core functionality, but I consider this big news, even if it seems to be something like a testnet. Working drivechains would be the perfect complement for LN and the main chain.

Blockstream's sidechain, called Liquid, also went live last month, and it's "L-BTC" sidechain token is supported and backed by 23 Bitcoin exchanges and services.

Does anyone have any opinions? I believe Bitcoin users will be moving away from main chain transactions for more speed and privacy, but move back to the main chain for storing for more security.


Title: Re: Sidechains - state of the art?
Post by: alko89 on October 16, 2018, 10:05:03 AM
It seems that the first test version of Paul Sztorc's Drivechain has been released at the end of September, without much news buzz:

http://www.drivechain.info/blog/first-release/

Usage tour: http://www.drivechain.info/blog/usage-tour/

I still have to investigate how they solved the problem of the missing Bitcoin Core functionality, but I consider this big news, even if it seems to be something like a testnet. Working drivechains would be the perfect complement for LN and the main chain.

Blockstream's sidechain, called Liquid, also went live last month, and it's "L-BTC" sidechain token is supported and backed by 23 Bitcoin exchanges and services.

Does anyone have any opinions? I believe Bitcoin users will be moving away from main chain transactions for more speed and privacy, but move back to the main chain for storing for more security.


I think thats what Lightning ntwork is for.


Title: Re: Sidechains - state of the art?
Post by: DooMAD on October 16, 2018, 10:50:30 AM
Blockstream's sidechain, called Liquid, also went live last month, and it's "L-BTC" sidechain token is supported and backed by 23 Bitcoin exchanges and services.

Does anyone have any opinions? I believe Bitcoin users will be moving away from main chain transactions for more speed and privacy, but move back to the main chain for storing for more security.

I think thats what Lightning ntwork is for.

Liquid and Lightning are designed to perform different roles.  There's a decent article here (https://thebitcoinmag.com/2018/10/15/bitcoin-sidechain-liquid-network-goes-live/) about the launch, which does a good job of highlighting Liquid's purpose.  The crucial distinction is that Lightning involves sending transactions off-chain, whereas Liquid is a pegged sidechain.  Liquid transactions are recorded on a blockchain, but not the same one as Bitcoin.  The companies listed in that article are basically maintaining their own blockchain.


Title: Re: Sidechains - state of the art?
Post by: d5000 on October 16, 2018, 02:33:57 PM
Blockstream's sidechain, called Liquid, also went live last month, and it's "L-BTC" sidechain token is supported and backed by 23 Bitcoin exchanges and services.

Does anyone have any opinions? I believe Bitcoin users will be moving away from main chain transactions for more speed and privacy, but move back to the main chain for storing for more security.
Blockstream's Liquid and Rootstock are federated sidechains - they are pretty trivial to implement as the peg and the blockchain maintainance is managed by a group of companies. The challenge is to find trustworthy members for the "federation" holding the multisig accounts.

However, Liquid/Rootstock may have interesting use cases - mainly for transfers between exchanges and service providers, which seems to be their main purpose.

The "Drivechain" model is more interesting for me as the peg in this case is maintained by miners, and thus is, in theory, decentralized. In the case most Bitcoin miners mine the sidechain too, the attack cost for the sidechain should be similar to the attack cost of Bitcoin's main chain itself.

I agree that - once transaction fees rise again - sidechains and LN could largely replace mainchain transactions for the general public, if they're implemented in a proper way, i.e. in an easy-to-use GUI client. LN would be used for microtransactions, sidechains for most mid-tier transactions and the mainchain only for the bigger ones (settlements, real estate investments etc.).

One thing about "sidechains" is that they don't really need to be blockchains. You could use a normal database and store hashes of that data on the mainchain. I think that is very efficient.
But wouldn't you need a centralized entity to inscribe the hash data in the main blockchain? Otherwise, there should be a double-spend risk with this approach, or you would have to inscribe each single transaction in the main chain, too ....


Title: Re: Sidechains - state of the art?
Post by: nicosey on October 16, 2018, 05:00:42 PM
You also have the Ocean sidechain from CommerceBlock

They attest to the BTC via mainstay https://www.commerceblock.com/wp-content/uploads/2018/03/commerceblock-mainstay-whitepaper.pdf


Title: Re: Sidechains - state of the art?
Post by: domob on October 16, 2018, 05:50:32 PM
You also have the Ocean sidechain from CommerceBlock

They attest to the BTC via mainstay https://www.commerceblock.com/wp-content/uploads/2018/03/commerceblock-mainstay-whitepaper.pdf

My understanding is that Mainstay is an alternative form of merge mining for securing altchains.  In this topic, however, the OP is asking specifically about sidechains whose coins are pegged to BTC.  Does the Ocean sidechain have a two-way peg with BTC (the whitepaper you referenced here doesn't talk about pegs at all)?  And if so, how is that implemented if not through a federation of multisig holders?


Title: Re: Sidechains - state of the art?
Post by: nicosey on October 16, 2018, 06:06:26 PM
Sorry I thought the OP was asking about different trust models.


Title: Re: Sidechains - state of the art?
Post by: d5000 on October 16, 2018, 07:08:52 PM
I'm indeed mainly interested in two-way-pegged sidechains without a "federation" securing it - i.e. with a "decentralized" mechanism like merged mining. But there is so few content about sidechains at Bitcointalk that we can handle everything in this thread - consensus and pegs. So @nicosey, thank you for the link to Mainstay.

I only read the first couple of pages but it seems interesting as it provides a method to ensure the federation members do not attack and fork the sidechain to steal tokens from the regular users.


Title: Re: Sidechains - state of the art?
Post by: nicosey on October 17, 2018, 06:53:28 AM
I only read the first couple of pages but it seems interesting as it provides a method to ensure the federation members do not attack and fork the sidechain to steal tokens from the regular users.

Correct, we thought that we be an approach that could be useful.  As we don't need permission from the miners to secure the chain.


Title: Re: Sidechains - state of the art?
Post by: d5000 on June 14, 2019, 06:13:23 AM
A little update about the Drivechain project (which I'm loosely following ... sort of ;) ):

  • The Git project (called "DriveNet") seems pretty active: see here (https://github.com/DriveNetTESTDRIVE/DriveNet) - last updated 2 days ago; last update is creating a new (testnet? sidechain?) genesis block, so they may be about to test new features.
  • This tweet (https://twitter.com/Truthcoin/status/1105572522961952769) from March is also interesting: he writes that Drivechain software is "almost finished", and gives out a link to a Telegram group (https://t.co/L3R0hqf7mL) for "DriveNet Insiders". Sidechains on the testnet can be created via a GUI.
  • Paul Sztorc has published a new blog article in April: Why Drivechain is harder to understand than previous softforks (http://www.drivechain.info/blog/hard-to-understand/). Unfortunately, it seems partly a bit of a rant against the Bitcoin Core project. That is an indication that it could be difficult for the code to be included. (Additionally, there is a little anti-Liquid rant (http://www.drivechain.info/blog/liquid/) in his blog, here technically however I agree with him.)
  • YouTube video (https://www.youtube.com/watch?v=Fc4knRshUKY) about Drivechain, already a bit older (Dec 2018) but newer than last update of this thread ...

In summary: I'm glad that this project continues, but it's a bit a pity that the relation between Sztorc and the Core devs seems to be a bit conflictive. I almost fear Drivechain will be deployed ... at an altcoin blockchain.