slow progress on properly handling hybrid orderbooks and at the same time eliminating any chance of losses, regardless of who generates the block, edge cases of when the transactions are confirmed, etc.
After finally understanding it better, I was able to get the code much, much simpler to where I could understand it.
So now to finish coding and testing it. Still need to deal with maintaining state as the event driven fill or kill model cant handle what is needed. But I think I found an intermediate step that isnt fully Tradebots supporting, but is halfway there. So I dont mind getting that working as it is in the right direction
Hope to have next progress report in about 48hrs
James
|
|
|
had a ramchain "bug" after investigating it seems was just a normal blockchain reorg of > 3 blocks this actually does happen, especially for coins with faster blocktimes
James
|
|
|
The Jay asset will be released in the near future!
We have all seen the amazingly cool stuff young Jones is making at jnxt.org. Jay will be the culmination of all this tech and it will allow for 100% lite wallet access to NXT and SuperNET. More specifically InstantDEX.
Yes, this means that people will be able to trade on InstantDEX without having to install any software, just download the HTML and run a localhost page to trade. InstantDEX has made a deal for a 20% revshare to Jay for all the InstantDEX trades it is part of. In the case where half the trade is via Jay, then it will be 10% revshare.
Jay is not a static thing and I expect over time, it will get more and more functionality and should become one of the primary paths for new users into SuperNET.
James
|
|
|
Sorry for a newbie question about assets withdrawals: I downloaded a Nxt client from nxt.org and announced my public key through http://jnxt.org/key/, so now I can withdraw UNITY and all related dividents of SuperNET from BTER to my NXT address? Thanks! you should be able to just do one at a time to make sure you are able to properly access your account James
|
|
|
The long winter is coming to an end.
I will be announcing an upcoming NXTventure dividend soon
James
|
|
|
Repost from SuperNET forum: Tried to test the new SupernetV1.1 beta. In the basic console I got "Error getting deposit address" errors (BTCD, Bitstar, Opal, Vericoin, VpnCoin). Any ideas? It seems like some sort of network error, but this is strange. I have asked for help to troubleshoot this as it is in the GUI code James
|
|
|
Anyone an idea here how long we have to wait before we can withdraw Supernet from bter. Or is this all a big joke and don't we ever see any coins back?
All the indications are that coins will be returned (not really sure about BTC, though), but they seem to be taking each coin or block of coins in turn, without indicating whos turn it is in the cue. For now, its twiddle thumbs time, and hope for the best. I expect next week NXT will be opened for withdraw and hopefully all the assets at the same time James
|
|
|
I think it makes sense to move direct to Tradebots ordermatching support. no sense to get another interim method working, where there is already the fill or kill order matching working
it is a bit surprising how much cascade effect allowing for partial fills creates...
to properly solve it requires a far more sophisticated approach than event driven fill or kill
dont worry if no progress report for a while, this will be a bit more than usual coding
James
|
|
|
A bug was found!
I limited the length of allowed strings to the size of a pointer (8 bytes), instead of the 10 bytes that was needed. But I have a feeling there are some more bugs to be found...
I seem to be making the new functions faster than they can be tested. Need to get some "breakfast" and onto the final InstantDEX loose ends
James
|
|
|
almost have hybrid books, actually it is coded, but it raised some secondary issues that need to be sorted
also need to extend the trade negotiation sequence and to allow for handling them in parallel.
At that point, it will also close the one attack vector that the block generator could do and InstantDEX core will be in maintenance mode for me along with ramchains and realtime MGW. Barring lots of bugs being found, I think I will be able to start finalizing Tradebots core next week.
I am also in parallel with all this revamping the low level networking to add defenses against various attacks. This is using tech from Jinn Labs, and after this SuperNET will have arguable the most secure and private network that everyone can use.
Then the final pieces of the SuperNET core, Teleport and Telpathy, can finally be activated. I did say it would take me about 6 months and having to do ramchains pushed the timing back a month, but overall I think the progress is not so bad.
I am seeing a lot of positive feedback from the v1 GUI, keep in mind I am just talking about the SuperNET core which is now around 75 API calls.
James
|
|
|
Added merging external orderbooks in, so you can see the prices on poloniex, bittrex, cryptsy, NXT AE in the same orderbook
|
|
|
It is the choice between cute blonde and sexy brunette... Clearly there is only one solution
James
|
|
|
still waiting for GUI release, but the Tui is used for testing and is fully connected to the InstantDEX API
I got the ordermatching working, along with order expiration, so it is theoretically functional now. Of course it needs to be tested first
James
|
|
|
Here is debug Tui by cryptosleuth: ![](https://ip.bitcointalk.org/?u=https%3A%2F%2Fi.imgur.com%2FhJkXiY0.png&t=663&c=Kg2FCYuSssODrA)
|
|
|
finally got the order matching to pass all the checks and do an order. it is quite a few checks that need to be passed also have multiple orderbooks autocreated so soon we can get a combined orderbook from many sources
James
|
|
|
@altcoinUK
InstantDEX is in testing now, it builds on the NXT AE that has been working for almost a year. So this is something that already exists. The InstantDEX is designed to enhance it to make it usable like a centralized exchange, in fact we are starting the first GUI using peatio which is open source centralized exchange software. Of course we are removing all the KYC and other centralized things, but the GUI is similar to most exchanges: yunbi.com/markets/nxtbtc
It seems that via different path's Anonymint and I have come to similar conclusions as what is needed, of course since I am not having formal educations, my words are using my meanings and I cant write whitepapers. Still this is the final missing piece that is needed for mass market adoption of crypto. And there will be more than one crypto that survives, so it makes perfect sense to have a framework for cryptos to connect with each other, either directly at the protocol layer or via indirect market mechanisms
Anyway, I still have a month or two of debugging to get SuperNET core into testing for all my major components, so until then this project is to be percolating.
James
|
|
|
I didn't read that whitepaper, but if I understand correctly how a DHT would be used in this scenario, it means everyone knows which IP addresses map to given hash in the DHT (because they need to know which IPs to send the copies of the encrypted message and only the destination IP knows how to decrypt it).
Just wanted to clarify this point. The assumption that the IP address is known is not necessarily true. What I do with Telepathy is send an encrypted packet with an additional encrypted payload and send it to a deaddrop address that belongs to no node and therefore no IP address. This deaddrop address belongs to nobody but it has a special property, it is within the n-dimensional R-ball so that by modifying the DHT routing logic a bit, it can be assured that the actual intended node routes the packet following the rules all nodes follow. There are no servers and all the nodes are peers. And each node has a public node address that is bound to an IP so the DHT routing works, but the private comms are to the deaddrop address that is close to the private address. As the network grows the number of possible destinations continues to grow (surface/volume ratio) It is true that with multiple communications sequences an intersection of online nodes that are within the R-ball can be created, but if the deaddrop address is shared within a neighborhood, this becomes like a public phone in a public square. The phone rings, but nobody answers it. Still the message is sent as only the intended destination is able decrypt the extra encrypted payload. Even in a network with a high percentage of attacker nodes, there is no definitive indication that any message was received. By making a new private address and a corresponding new deaddrop address, correlations over time are prevented. On the sending side, it is possible to inject a DHT from a random node send via onion packet, so the origin of a telepathic packet has the entire network as the anon set. By breaking the link between IP address and destination node, not even the sender knows the IP address of the destination, nor do any of the other nodes that do the routing of the DHT. James
|
|
|
This is a text based GUI for InstantDEX API that we will use for testing: https://i.imgur.com/rdbR52W.jpghttps://i.imgur.com/NkjOZ5N.pngnothing fancy, but will allow for realtime order entry and trading. So we are going full speed ahead with the InstantDEX testing and hope to have the low level rock solid for when the graphical user interface is ready James nice, everything is starting to fall in place now ![Smiley](https://bitcointalk.org/Smileys/default/smiley.gif) ) not long until a nice looking gui is on the way i guess? if possible maybe something like https://beta.bittrex.com/ ; it is down right now, but looks really good and alot of extra information! it would require someone that knows code to verify, but it was very small changes to reactivate InstantDEX as it has been 90% complete since summer. just so many things happening... anyway i am the turtle. zoom zoom James
|
|
|
This is a text based GUI for InstantDEX API that we will use for testing: https://i.imgur.com/rdbR52W.jpghttps://i.imgur.com/NkjOZ5N.pngnothing fancy, but will allow for realtime order entry and trading. So we are going full speed ahead with the InstantDEX testing and hope to have the low level rock solid for when the graphical user interface is ready James nice, everything is starting to fall in place now ![Smiley](https://bitcointalk.org/Smileys/default/smiley.gif) ) not long until a nice looking gui is on the way i guess? if possible maybe something like https://beta.bittrex.com/ ; it is down right now, but looks really good and alot of extra information! GUI is based on yunbi.com/markets/nxtbtc, we are removing the centralized stuff. I wanted a GUI that is familiar to users. The v1 GUI is in very active testing now, but that is just the first part of the iceberg as the InstantDEX is developed independent and then it gets merged in. Along with all the other projects, so it is confusing to those that want to buy the rumor and sell the news as there is always things happening James
|
|
|
|