[NXT] Testnet for Automated Transactions (AT) - Turing Completeness for Nxt

[Momimaus]

Automated Transactions was an attempt to implement ideas similar to those discussed in early Ethereum discussions.

Ian Knowles and a Greek named Vasilis issued an asset called NxtAT that was somehow related to the production of these features.

NxtAT seems to have disappeared.  http://is.gd/mxGIU7

the current capitalization of NxtAT is roughly $486,000 USD.

the issuer account has, at the time of writing, a balance of 1,368,473.02 NXT.  https://nxtblocks.info/#section/accountId/NXT-RZ9H-H2XD-WTR3-B4WN2

I really don't have a dog in this fight, but some people do still seem to be developing Automated Transactions.

Any statements from Ian Knowles regarding this?

Give the money back thief!!!!

[fivebells]

You might be interested in the machine code testing and results for Nxt AT: http://ciyam.org/nxt/

Thank you, that is extremely helpful.

[Daedelus]

You might be interested in the machine code testing and results for Nxt AT: http://ciyam.org/nxt/

Thank you, that is extremely helpful.

You are welcome  I have to say it is nice to have a civil discussion on BitcoinTrashTalk for a change.


You may also be interested in Smart Contracts by Come-from-Beyond >>> https://nxtforum.org/smart-contracts/express-your-wish-(smart-contracts)/

It's not Turing Complete but some devs argue that TC is OTT for most applications. This is a planned core-feature of Nxt but clashes with Nxt AT, which has created friction between the two projects.


Nxt also has a third implementation, "Community-checked Programmable Agreements" (catchy, yeah?  ), previously named Semi-Automated Transactions >>> https://nxtforum.org/automated-transactions/semi-automated-transactions/

Again, not Turing complete but the argument in there covers why. Too technical for me but you might like it.



Why have so many of the same implementation? My understanding is that Smart Contracts was always a planned Nxt feature. Nxt AT was started by a third party dev, not knowing about Smart Contracts as it was unannounced at the time so there is overlap. Kushti is a part time core dev and his project is done as a third party dev. Third party devs can build what they like so if they see value in it, who are we to judge? Whatever happens, it is all still exciting



TL:DR Nxt has three implementations of Smart Contracts in development

i) Nxt AT (Turing Complete)
ii) Smart Contracts
iii) Community-checked Programmable Agreements (previously Semi-Automated Transactions)

[fivebells]

Wow, awesome stuff, thanks again!

TC might be a bad idea, anyway.  Seems like any sufficiently powerful cryptocurrency may contain the seeds of its own destruction.

[Daedelus]

I think I know what you mean. There has been some thought to the problem of malicious scripts in Nxt AT and I understand the devs believe that they can contain the threat. I don't know what is discussed behind the scenes/devchat but publicly they plan to:

I) limit the access the Nxt AT Account to a subset of the NXT APIs and;
II) each cycle of the Nxt AT contract has a cost in Nxt so it can only run until the account runs out of money. It can't run forever and bloat the blockchain, an attacker would have to keep buying and sending Nxt to the malicious AT to keep it attacking.


Interestingly, L8rre argues here >>> https://nxtforum.org/automated-transactions/ats-with-fsm-based-dsl/msg30133/#msg30133

...that Turing Complete can't be Turing Complete as it can't calculate 1+Pi, it could only run forever and ever. Nxt AT above would be able to calculate this to as many decimal places as the Nxt in the account allowed. So is this really Turing Complete?   But I think Kushti clears up the definition later on down the thread (also note how Nxt forum tolerates clueless noobs like me, I learn a lot  ).

[Daedelus]

I realised there is another Turing Complete implementation in Nxt done by a third party developer.

It is for a specific subset of tasks and not meant to be as far reaching as Nxt AT AFAIK, I'll dig out the link...

[superresistant]

 
This is epic. Could Nxt be the first ?

[achimsmile]

Watching. I would love to see a decentralized version of kickstarter 

[Daedelus]

This is epic. Could Nxt be the first ?

Yes, I believe Nxt is the first to have a working product implemented (for historical accuracy, CIYAM is the inventor and made the first version of AT in C++ so that was first. Though I don't think it was ever 'in the wild' with public access. Nxt's implementation is in Java).

Look at the dates of the podcast, the OP and you'll see it has been in development for over 6 months. The testing from the last 3 months published in the link above. Soon it will be released for the public to play around with (and for more testing with people trying to break it  ).



The developers of Nxt AT are asking for bounties to implement AT in other coins so I imagine it will be similar to the Asset Exchange: Nxt will get the headstart but there will probably be others coming down the track*

The bonus of this is that you will be able to do "Atomic Cross Chain Transactions" between different coins that have AT implemented. This boils down to the AT acting as a trustless escrow, send your Nxt for another coin at a set price to the AT. If the other side doesn't send their side of the deal within x blocks, the AT will refund you. All decentralised and on two blockchains.

Or maybe the AT straddles both platforms with standing balances in both, so you would be trading directly with a algo driven robot who puts up the bids and asks   I'm not sure of the implementation they have in mind.


I don't see why Smart Contracts or "Community Agreed..." couldn't do this use case (though don't quote me!) but Nxt AT contracts are Turing Complete. And Turing Complete contracts is what got people wetting their pants about in Ethereum.


 *(Nxt Asset Exchange trade amounts to $2.2 Million USD in the last 30 days at current low exchange rates, I'm not aware of another decentralised exchange comes close to that. Daily trade volume on the Asset Exchange is reasonably correlated and regularly exceeds that on Coinmarketcap, as it does today. Source: http://nxtreporting.com/assetcap.php)

[Brangdon]

I think XCRs will see more developer use and support because of it's low cap high potential aswell as it hasn't dwelled into every bit of business and copied every idea out there in crypto there is, also can build what nxt build in it and vice versa. So I think both will find a nische in the market with XCR winning as it is the core function intended which will get most focus.

The cost of attacking a coin generally depends on its market cap. For example, the more BTC is worth the more the block reward and transaction fees are worth, the more hash-power the network gets and the more expensive it becomes to acquire 51%. Similarly with PoS coins, the higher the market cap the more expensive buying 51% of stake is. So, if you want to build a feature on top of a block-chain, and you care about security, pick the one with the highest market cap. Last time I checked, Nxt's market cap was 60 times XCR's.

(I'm not getting at XCR especially. This argument gets used against Nxt, too. Some say everything should be built on top of Bitcoin because its market cap is so high.)

[bluemeanie1]

Has Ian Knowles actually made any statements lately regarding NxtAT?

Is it still alive?

[achimsmile]

Has Ian Knowles actually made any statements lately regarding NxtAT?

Is it still alive?

 
www.nxtscam.org

cute

[fivebells]

Has Ian Knowles actually made any statements lately regarding NxtAT?

Is it still alive?

Some time ago, Knowles said he's abandoning it due to disinterest from NXT core developers.  However, if you follow the link I gave above you'll see that NxtAT is still very much alive.  Presumably Knowles gave the funds to the developers who are continuing the project.  You might ask them for verification.

[Daedelus]

Some time ago, Knowles said he's abandoning it due to disinterest from NXT core developers.  However, if you follow the link I gave above you'll see that NxtAT is still very much alive.  Presumably Knowles gave the funds to the developers who are continuing the project.  You might ask them for verification.

Ian has left NxtAt (hot off the press yesterday) presumably to implement AT in the first coin to meet the bounty. Vbecas and btc2nxt are developing NxtAT.


I don't usually acknowledge Bluemeanie but as he keeps intruding and adding nothing...


BM knows the state of AT, he follows nxtforum very closely. He is just trying to undermine it. He takes Nxt's positives and then tries to pillory them: Nxt itself > Jl777 > Supernet > Threatening Nxters with legal > Doxxing single mothers, trying to get them fired and now it looks like CIYAM is next.

I wouldn't recommend trying to enter a dialogue with him, he has a history of ignoring my posts and I haven't seen him answer a question yet. Or provide any evidence for anything he claims, working on the basis that "throw enough sh!t some of it will stick" rather than "this is my claim supported by evidence a, b, c...".

This is what he is trying to obfuscate > http://cointelegraph.com/news/112643/the-mystery-of-the-missing-1000000-nxt
BM acts as though it is an opinion piece but the blockchain doesn't lie and his promo messages are signed with his PGP signature. You can make up your own mind, everyone who has looked at the evidence is unanimous in their verdict.


Fortunately, this mud slinging tactic of his will only have an effect on those who don't do their own research. So you will be fine   I haven't made it onto his 'enemies' list yet but it is only a matter of time. After all, I have publicly disagreed with him

[itsAj]

Automated transactions for NXT could be huge!

[Daedelus]

I think it will be a slow burn to begin with but with it being implemented in other platforms, the user base could grow quickly.


I mentioned there was another Turing Complete implementation in Nxt:

NXT has Turing complete language via Tradebots (formerly NXTcoinsco) service!

https://nxtforum.org/general-discussion/price-speculation/msg71681/#msg71681


Come-from-Beyond mentions that there is overlap between this and his Smart Contracts on the following page. This one would be hard to find if you didn't know it already existed, buried in a megathread.


So, Nxt's smart contract implementation:


i) Nxt AT (Turing Complete)
ii) Smart Contracts
iii) Community-checked Programmable Agreements (previously Semi-Automated Transactions)
iv) Nxt Tradebots (Turing Complete, previously named NxtCoinsCo)

[CIYAM]

Just to clarify a couple of points:

I returned all the NxtAT Asset units that had been sent to my account to the account that issued the Asset (which was never mine) as well as all my NXT (so in fact for the > 100 hours worth of work that I did I have kept only some BTC from private donators as I returned all the NXT that was *sent to me for my efforts helping both Vasilis and Adam*).

To be very clear - I no longer *do any work on the Nxt AT* implementation of AT (and I barely made more than about 3 BTC from the whole thing).

The current AT documentation can be found here: http://ciyam.org/at and you'll note I have been working on how to get AT to work with Bitcoin/Litecoin clones as well.

Please note that also *it was my invention* (and is coded in C++ *not* Java) and it has always been "open source".

[Daedelus]

I have strengthened my post in response to the "first implementation in Nxt?" above to make clear the credit you are due as being the inventor of AT.

I am personally grateful to you as one of my 'mentors' (willingly or not) that took me from crypto noob to not-so-crypto noob over the first 6 months or so at the start of the year. You always had time for my questioning, always dumbing down and rephrasing when needed  A lot of people are only just hearing "atomic cross chain transactions" but we have already been through a few rounds of description already for me to 'get it' (almost enough to put it in the glossary without fear of being wrong)  I am sad to see you go and hope in time there will be no lingering hard feelings between Nxt and your next projects.

[Daedelus]

Notice to interested Devs:


vbecas plans to release the code for NxtAT today for review.

He doesn't plan to run it on a testnet until the weekend, so that he can provide support for testers as he will be at home to deal with any error.


https://nxtforum.org/automated-transactions/information-of-the-at-project/msg117316/#msg117316

[Daedelus]

NxtAT needs node to test, testport = 5876, peer =5874
anyone can help?

If you want to take part in the testnet, please go here >>> https://nxtforum.org/automated-transactions/information-of-the-at-project/msg117224/#msg117224