Bitcoin Forum

I would like to announce a release (http://www.coincer.org/2015/09/14/first-release/) of the first version of my decentralized cryptocurrency exchange project – Coincer (http://www.coincer.org/). Although version 0.0.1 is still more like an unstable preview running on testnets, it's already functional.

Coincer is decentralized by means of TierNolan's atomic protocol (http://www.coincer.org/2015/02/06/atomic-protocol-3-final/) (just a bit enhanced). Announced version 0.0.1 still uses a central server to forward communication between users, but this is just a temporary solution. I'm already working on a P2P overlay that will fully replace it. This is something that no other similar project has offered so far.

Currently there is only a CLI available (see a demo (http://www.coincer.org/2015/09/14/first-release/)), but GUI is of course planned too. Nonetheless, it's not that bad as you might think. ;c) Coincer doesn't connect to any cryptocurrency network itself, but relies on other clients (currently bitcoind-based RPC is supported). I also plan on publishing detailed specifications so that anyone could easily create an alternative client compatible with the rest of the network. Developing a network and protocols is my primary goal.

I will be grateful for your feedback. If you feel like supporting the development, I will be glad too. :c) My bitcoin address: 1DmvwcYUkRUbDdfe2dKVEkZJAY7Bwv9zy9

And the most important link: Download (http://www.coincer.org/download/)

---

Current version
Download version 0.0.5 (http://www.coincer.org/download/)

Changes in 0.0.5:

• encrypting all trading messages (using ChaCha20 stream cipher and Curve25519 keys)
• automatic market exploration on startup
• running also as a daemon
• routing enhancements
• switched from hop-based to time-based metric for building routes
• mitigation of replay attacks

 Thank you great project,will download documents read,and will ask you some questions if it is needed,i need exchange for currency what is not blockchain,but has rpc cliant done,custom code can be done if needed

Nice project you have completed here mate, have you checked it for bugs, backdoors and likes?
I am also interested in the virus scan of the script and it will be great if you can post the result here.

Yes, there definitely are bugs in this version (0.0.1), I know about several of them, that's part of the reason why it has such version number and is called unstable. ;) Some of them will get fixed by further development, the rest will need to be fixed explicitely. But I can assure you that I'm not going to release any stable release that would still contain them. :)

Ok, that is cordial of you, I will sure like to use this when it become more stable, I have bookmark this page in order to watch the development on this thread.

Is this Still being developed? Nice project by the way.

Sure, I'm working on it (currently on communication over P2P). :) Thanks!

Very nice. I will watch this thread for progress. Would you make some demo videos?

Version 0.0.2 is out. (http://www.coincer.org/download/) Its main feature is Coincer's own P2P overlay (http://www.coincer.org/2015/10/08/draft-of-p2p-overlay-design/) so that no central server is needed anymore. There is just a seed DNS server that serves as a source of node addresses to bootstrap new peers. I run one node on the server that I also use a source for the seed. Please note that there isn't any active market discovery yet, just a passive one.

This release was delayed for couple of weeks, because I didn't know how bureaucratic process getting a permission to run such a basic thing like a DNS server would be. :-[

Looks very promising. I can hardly wait for decentralized exchanges become operational. Bookmarking this thread...

Rolling out version 0.0.3. :D It contains several improvements and bug fixes in the P2P code (discovered during testing of previous version), adds logging (useful thing to know what is happening on the background, while not flooding the terminal), and a command to print current peers (Coincer itself automatically tries to keep at least 4 connections to the network), and few minor things.

Until now the problem was mainly in transaction malleability (which breaks rollback transactions), but as this is finally being fixed, I'm very optimistic about operationability of this and similar decentralized exchanges. :)

Looks good, hope you have the right infrastructure in place.

Version 0.0.4 is released. From the most notable changes: nodes now maintain their local databases of peers' addresses so that the DNS seed is now a secondary source of addresses (used only if there isn't enough of live nodes in the list); and all P2P messages are now being signed. I chose Ed25519 as the signing algorithm and already made some steps so that the same key could be also used for encryption. As nodes use their public key as an identifier in every message, it's really easy to do verifying of signatures and later also encrypting as there is no need for additional key exchange. :)

Today I released new version 0.0.5. It adds messages encryption next to the signing and thanks to ECC, there's no need for extra key exchange mechanisms – everything is already built in. Another nice feature is automatic market exploration that makes it possible to join the network and to almost immediately see what is there, what are the active orders. Aside from various improvements there is also a protection against DoS attack on routing and a protection against replay attack.

This Sunday, 13th December, I'm going to hold a testing&trading session starting at 15:00 UTC.

Please make sure you have at least two out of three currently supported testnet clients (BTC/LTC/DOGE Core) synced and that you can run Coincer. Nonetheless, only one sort of testnet coins is needed as you will have chance to exchange them for different testnet coins. ;)

I will prepare a special patch for this session that would create a log with detailed information on functioning of Coincer, its P2P network, or trading. Sending this log back to me will help me a lot in further development.

Primary communication channel during this session will be the official IRC channel (#coincer at Freenode.net).

You can still jump in! If you don't have any testnet coins, I can provide you with some. :D Thanks a lot!

Seems like a good concept.
Can't wait till it is working, might try it out later.

Thank you everyone who took part in the test! :) We found several bugs and I gathered some data about the network to analyse. I would also like to thank a lot Dogecoin devs who fixed the testnet chain for me. 8)

I promise that the next testing session will be annonced a bit more ahead. ;D

It seems like a pretty useful development. So... will we still see this coming out?

Absolutely! :)

whats the status on this?

Very nice project ! Good luck :)

Oh, you just remined me that I haven't finished my summarizing post on this. Anyway, I don't have much time this nor next month, so there won't be probably anything new happening. Then (in March), I'm going to focus on improving P2P layer, communication protocol and probably also refactoring trading, because I need to extend it a little bit, but its current state isn't very nice. You can also expect another testing session. :c)

Thanks! :)

Seems we are doing very similar things: https://bitcointalk.org/index.php?topic=1340621.0 (also based upon TierNolan's basic idea but reworked to use CLTV with P2SH so no malleability issues and txs should be considered as "standard" for both Bitcoin and Litecoin now).

If you would like to use some of the CIYAM code to get your implementation using CLTV and P2SH you would be welcome to (also feel free to PM me if you wish to discuss it).

Also if you are doing a GUI then with some very minor tweaks you could get it to work with AT blockchains as well (Qora and Burst) as they have supported ACCT for a long time now.

CLTV is one of the things I have in mind when I write about extending the trading part of Coincer. :) On the other hand, I'm not going to completely abandon current approach. But maybe you know more than I do about current malleability issues in Bitcoin? Because as far as I (think I) know, there shouldn't be anything that could affect this atomic protocol. With BIP66, all parts of BIP62 should be fixed (except those where the malleability comes from ECDSA signatures, but those aren't relevant here as it's the sender who strives for the txid remaining the same and no one else could be affected).

I'll definitely have a look at both Qora and Burst. :c) Although a GUI has a low priority for me, it still is on my ToDo list and eventually will be implemented.

I think the main issue would be if the forked alt has not updated its tx signing code (as the sender's txid could be changed without their knowledge).

By this I am referring to the low S values issue which I patched for CIYAM here: https://github.com/ciyam/ciyam/blob/master/src/crypto_keys.cpp#L531

Other than that I think that using P2SH (rather than the original multisig) is just a neater approach (it probably doesn't alter the workflow though).

Yes, I try to keep this in mind. Before including an explicit support for an altcoin, I'll have to make sure that it meets certain requirements (like supporting CLTV or non-malleability or possibly something else). Moreover, there also might be altcoins that use different hash functions or generally script encoding which should also be taken into account.

But honestly, I don't expect too many of them to include needed features. ;)

xhire ...

what a great project ...

why dont you add it to git - so those who are interested could help ;) ...

#crysx

Thanks! :)

Well, if you go to Download (http://www.coincer.org/download/), there's a link under "Source Code" leading right to the git repository (https://bitbucket.org/xHire/coincer-ruby)! :D

kool - and tanx ...

bitbucket ... i use github ...

but nonetheless - this is a very good looking project ...

tanx mate ...

#crysx

Of course, I'm a fan of decentralisation. ;D

hahaha ... and you are a funny bugga ...

nice one ...

im off to bed now ... cant wait to see what the working product might look like ...

#crysx

after the fall of cryptsy - decentralized exchanges are the futur.
Bice project

Very interesting project. Assuming the platform is stable can it be a safe place to store coins as well as trade?

Well, some people suggested that it might be good to implement simple wallets for the coins and connect directly to the coin's networks, but it seems to me to be needlessly complicated. So far I prefer to connect to other clients, that provide wallet functionality, via their API. Currently, I already have some support for bitcoind-like API and would like to add support for Electrum and possibly other clients as well.

It should be pretty safe way of trading, although probably not the cheapest and definitely not the fastest (because it's happening on blockchains so one has to wait for confirmations).

I like the project but the name, not so much. Following....