Pages:
Author

Topic: [ANN] Coinbend: a functional decentralized cryptocurrency exchange (Read 3945 times)

full member
Activity: 179
Merit: 250
Everybody love the concept but all of the similar porjects never go anywhere. I think that this is going to end up being a niche field for doing very large transfers between to parties.
legendary
Activity: 2352
Merit: 1024
Leading Crypto Sports Betting & Casino Platform
I found this thread by doing a google search for "p2p coin exchange"
I've just begun to read some of your resources.
How are you making out with this?

I've stopped working on it due to a complete lack of interest so the project is dead at this point.


wow unbelievable this lack of interest from the community. this project looks like what crypto needs atm, esp after the drama with cryptsy now. sending the link to this thread to some friends who might be able to help with some stuff. best of luck with the project!
member
Activity: 114
Merit: 16
Quote
This isn't that important at an early stage of the software.

It's important if you want people to use it. One of the main reasons why people keep putting their coins into centralized exchanges is because its easy. Who wants to install a bunch of clients manually and then wait days for their blockchains to download? Nobody.

Quote
This is probably the most important part and should get most of the attention.

Agreed, but easier said then done. I would also add a better way to test the software with simulated blockchains. Could work as a fake JSON-RPC API for the coins to simulate transactions being confirmed as that's mostly what the software uses for validation.

Quote
I agree with you that security is important. But there are also other important features that an exchange must have and it seems to me that a timechain may inhibit one important feature and that is speed. Now maybe I am reading/understanding timechains incorrectly...and I apologize if I am. But when I see a "5 minute interval" I'm understanding that transactions (trades) can take up to 10 minutes ... Am I correct in this thinking? If that is the case then I don't recommend taking this route as the software will be lacking an all too important feature.p

The main benefit that a p2p exchange can provide is security and the current model is largely flawed without a solution for the trust problem. Timechains are that solution as they allow two unrelated systems to be linked with standard smart contracts without relying on a human for fail-safe refunds. Especially important given that cryptocurrency exchanges are still getting hacked.

Quote
In short: IMHO, If this is to work, Coinbend needs to have all the features of a centralized exchange + p2p + security

It's possible but in my opinion: not without significant funding. Nobody has the time or money to educate every potential user on a completely new way to transact + write the thousands of lines of code needed to make that easy. My expectation is that software like this will eventually emerge but it probably won't be trustless or user-friendly causing it to be niche and largely useless. The true challenges to solve for a project like this are actually immense because assuming you write all the code and the software works flawlessly: you still have to contend with the network effect which basically states that for certain applications the usefulness of the service is defined by the square of the number of users. So take Coinbase: why are they so popular? They have a huge number of users and thus a high liquidity; thus a higher potential to gather new users which just keeps on going and going. But how did they start initially and break through the noise of the online world? They had significant VC backing to the tune of millions. So you see, this project is much more difficult than producing the correct algorithms, and all of these problems add up to it not being worth it.
hero member
Activity: 868
Merit: 503
Are you excepting new coins?
sr. member
Activity: 367
Merit: 250
I'm a miner not a minor.
Some opinions and things that i've been mulling over...

It is my intention to provoke thought and possibly start a conversation as well as to help me (and possibly others) understand a few things. That is all Smiley

Quote
* Add thin-clients for the coins so full blockchains don't need to be downloaded (maybe using a consensus algorithm on top of the blockchains)
This isn't that important at an early stage of the software.

Quote
* Make the UI easier to use
I watched your video. I'm sure it could be more intuitive but it's a good start.

Quote
* Improve the trading algorithm to use a network-wide, ripple style consensus algorithm (to avoid front running and do proper p2p matching, current algorithm is crappy and may have problems)
This is probably the most important part and should get most of the attention.

Quote
* Timechains could also be removed but I honestly wouldn't recommend this if you give a damn about security. No timechains = trust and trust is a disaster waiting to happen when it comes to p2p finance and digital asset security.
I agree with you that security is important. But there are also other important features that an exchange must have and it seems to me that a timechain may inhibit one important feature and that is speed. Now maybe I am reading/understanding timechains incorrectly...and I apologize if I am. But when I see a "5 minute interval" I'm understanding that transactions (trades) can take up to 10 minutes ... Am I correct in this thinking? If that is the case then I don't recommend taking this route as the software will be lacking an all too important feature.

In short: IMHO, If this is to work, Coinbend needs to have all the features of a centralized exchange + p2p + security
member
Activity: 114
Merit: 16
I found this thread by doing a google search for "p2p coin exchange"
I've just begun to read some of your resources.
How are you making out with this?

I've stopped working on it due to a complete lack of interest so the project is dead at this point.

If anyone wants to pick things up again the code is currently working but there's a few things left to do:
* Add thin-clients for the coins so full blockchains don't need to be downloaded (maybe using a consensus algorithm on top of the blockchains)
* Make the UI easier to use
* Improve the trading algorithm to use a network-wide, ripple style consensus algorithm (to avoid front running and do proper p2p matching, current algorithm is crappy and may have problems)
* Timechains could also be removed but I honestly wouldn't recommend this if you give a damn about security. No timechains = trust and trust is a disaster waiting to happen when it comes to p2p finance and digital asset security.

In short: it's still a lot of work to get this usable and since the community doesn't actually care about security there's literally no incentive for anyone to work on this. Not even purely because its a good idea.
sr. member
Activity: 367
Merit: 250
I'm a miner not a minor.
I found this thread by doing a google search for "p2p coin exchange"
I've just begun to read some of your resources.
How are you making out with this?
member
Activity: 114
Merit: 16
very interesting it would be possible swap from any asset - blockchain even a sidechain made
like the ones in the factom project ?

I'm not an expert on FACTOM - but possibly. The algorithm should work with any divisible assets but at the moment I'm only focusing on cryptocurrencies.
legendary
Activity: 1188
Merit: 1001
very interesting it would be possible swap from any asset - blockchain even a sidechain made
like the ones in the factom project ?
member
Activity: 114
Merit: 16
Quote
OK it is good to know that you take that into considerations because there are people that always try to out smart another people.

Definitely. Security is why I'm building the software.

Regarding Coinbend: are there any features you guys can think of that -aren't- related to currency exchange? I was thinking that it has some robust P2P code and it will probably be a while before this software is user-friendly, so why not add something simple to get the network started.
sr. member
Activity: 434
Merit: 250
No doubt this is a pretty idea but the general problem with any P2P is one person trying to play smart than the other, what security measure do you put in place to check this type of ugly situation.

It's simple really: time.

Most of the security measures in Coinbend are all in some way a function of time. You have incremental transfers of coins that take place over a time frame and you have time-based refunds. Both of these approaches allow the risk for trading to be deterministically reduced - even between complete strangers - no third-party needed.

It's an interesting approach. Perhaps one that might eventually lead to the creation of high security databases that would make large-scale data breaches impossible. (Such technology doesn't exist yet but it could if we wanted it to.)
OK it is good to know that you take that into considerations because there are people that always try to out smart another people.
member
Activity: 114
Merit: 16
No doubt this is a pretty idea but the general problem with any P2P is one person trying to play smart than the other, what security measure do you put in place to check this type of ugly situation.

It's simple really: time.

Most of the security measures in Coinbend are all in some way a function of time. You have incremental transfers of coins that take place over a time frame and you have time-based refunds. Both of these approaches allow the risk for trading to be deterministically reduced - even between complete strangers - no third-party needed.

It's an interesting approach. Perhaps one that might eventually lead to the creation of high security databases that would make large-scale data breaches impossible. (Such technology doesn't exist yet but it could if we wanted it to.)
sr. member
Activity: 434
Merit: 250
No doubt this is a pretty idea but the general problem with any P2P is one person trying to play smart than the other, what security measure do you put in place to check this type of ugly situation.
hero member
Activity: 633
Merit: 768
BTC⇆⚡⇄BTC
[...]
How it works
[...]
Since micro-transfer amounts are designed to be as small as possible: any potential losses that might occur will be negligible and the impact of a hacked exchange would have minimal effects. In order to further reduce the risk of p2p trading: Coinbend also includes a secure refund feature. Refunds ensure that an owner's coins will eventually return to them in the event of a broken trade and work by using a new cryptographic construct called the timechain instead of traditional malleable refunds.
[...]

Nice to know that you created some kind of refund feature and found a way to implement it into a decentralized exchange. I guess implementing that idea took a little time and some additional effort, didn't it?

[...]
(Also, please let me know if I should keep working on this as I've already put in a huge amount of time with no reward and a lot more is still needed to get this to beta.)
[...]
Sure you should carry on!

Congratulations for your great project.

P.s. don't forget to publish some account for receiving donations.

Keep up the good work.
member
Activity: 114
Merit: 16
This looks interesting.

Is there anywhere we can read a little more about the details of the smart contract? Is it simply this https://en.bitcoin.it/wiki/Atomic_cross-chain_trading

The smart contract is a modified micro-payment channel but the magic is in the way that it's been implemented to minimize trust and work across cryptocurrencies.
sr. member
Activity: 261
Merit: 518
This looks interesting.

Is there anywhere we can read a little more about the details of the smart contract? Is it simply this https://en.bitcoin.it/wiki/Atomic_cross-chain_trading
member
Activity: 114
Merit: 16
I've just added basic instructions for developers on Github including a vagrant environment that handles installing all the dependencies and everything needed for development. Should hopefully make working on this easier.

https://github.com/robertsdotpm/coinbend

It would be interesting if you released a demo on YouTube with comments to show how make the exchange. Any projects in the field of decentralized exchange is certainly interesting. Thanks!

That's a very good idea. I did kind of rush the videos. I'll do a proper screen cast / demo explaining the software soon.

In other news: I'm still looking for a partner who wants to help me take this to beta. If you like cryptocurrencies and want to help people trade without losing their coins to hackers - send me a PM  Cheesy
legendary
Activity: 2156
Merit: 1132
It would be interesting if you released a demo on YouTube with comments to show how make the exchange. Any projects in the field of decentralized exchange is certainly interesting. Thanks!
member
Activity: 114
Merit: 16
I have answered  you on pm

He is great coder and programer,also he has great knowledge from area of finance and economy,investments
if you will need web designer that we know one,if he will not has time for it,i can try to contact with woman with good knowledge about web building,designe etc

Awesome. I hope this goes somewhere

Quote
Very nice project you have there!

Will your software be available for others to use to start their own exchange?

It's a p2p exchange so users download the software and run it on their own computers to trade crypto-assets that they singularly control. In this way, there's no limitations on what assets can be traded and deposits aren't required making it much more secure than giving away your coins to trade on a centralized exchange.

I suppose you actually could use the software to allow anyone to setup their own exchange. Under that model, operators would run the software and accept deposits on behalf of their customers who could then trade currencies with other customers on that node in real time - quite an interesting thought because you could argue that's definitely one definition of  "decentralized exchange" and the risks for trading would still be reduced by using smart contracts, however,  that's not exactly what I had in mind for the software.

Still, if enough people want such a feature I can always make this a multi-user system. What's cool about this software is the UI is web-based and the code is Python - so anyone can run this software and access it from any platform with a web browser. The web is extremely flexible and accessible, so building the software in this way is like giving people an entire exchange in a box. Prettty cool. Anyway, back to work  Grin
legendary
Activity: 1596
Merit: 1027
Very nice project you have there!

Will your software be available for others to use to start their own exchange?
Pages:
Jump to: