I think, I also decided not to invest - except if the DEV answer my questions about real time hardware control and outlines how that will be possible at all with the Blackhalo solution?
I am mainly interested in the IoT offering, but it seems to me the IoT concept does not even exists on conceptual level. If there is not even conceptual solution, I can't see how this project will release a working, blockchain enabled, peer to peer Internet of Things software that controls hardware in real time.
I believe I already answered this, im not able to monitor this thread constantly because I'm usually working. But yeah you can do zero confirmations with multisig because your device or you hold the raw transaction. It depends on the application/hardware. Best that you give me some example or use case and I can explain to you how I would approach it.
This isnt dissimilar from NightTrader multisig engine which I have made. I just dont have any front end for it. Controlling hardware in real time is fairly easy when you know multiple signatures are involved. Add this to checklocktimeverify and you can have outputs convert to different things based on blocks. Handling raw transactions where you hold the other signature can leave you secure with the fact that you can do zero confirmations.
Thanks for your reply.
I am familiar with the green address concept of Bitcoin that requires a third party and allows instant transactions and I believe you are not referring to that. So if your solution conceptually is not that, let see a concrete use case: open a gate in a sport venue where users use their wallet to enter the venue and the gate controller device charges 10 BitBay coin for opening the gate. My understanding is, prior to opening the door the device must know if the 10 BitBay Coin is available for the buyer. The device must ensure the balance is available, it's not a double spending and it's a legitimate transaction, otherwise the door cannot be opened. Once the gate controller device is satisfied the balance is available, and only in that case it opens the gate. How long it takes for the device to verify if there is a sufficient balance on the address, because prior verifying the balance the gate cannot be opened? In other words, how long the user must stand for waiting the device to perform the service?
If you say so you have worked this out and willing to explain this, I am sure that's the case, and you will have a working solution :-))
One working solution is to have your funds locked before you get to the gate ... Thus your primary account would have locked funds set aside for public events and interaction with IOT ... So they cant hold your money hostage because the funds are set to convert back to a regular account on the expiration of the lock ... The gate only signs with parties that were confirmed in advance to be honest. Those parties would hold a double deposit contract with the manufacturer too.
All due respect David, this solution isn't practical and such implementation would make the use case completely uncompetitive to existing FIAT or even digital currency payment solutions. Currently payment systems, even the Bitcoin network with green addresses works in ad-hoc manner: you pay at the time of the purchase. In order to buy a pizza, hotel room, concert ticket there is no requirement to lock the fund to the seller prior to the transaction, not even with Bitcoin. I think such system that you have suggested wont be adopted at all.
I understand your solution would be still anonymous and decentralized, but having so much complication to make a payment to IoT devices would make the solution impractical in my opinion.
This starts boiling down more to logistics.
I disagree. This is not a logistical but a conceptual issue. What you have described is a workaround by asking users to comply with your work around in order to have the luxury of involving with a very-very complicated process.
Anyway, thanks for answering, apart from the IoT part, your solution will be great and I have no doubt you will roll out an excellent smart contract system. I suggest don't involve the IoT it at all. Firstly people have no idea what is it so it doesn't mean anything :-)) and no one is investing in this because the IoT feature, but once someone take a closer look at it, the reaction will be that it is not acceptable for businesses and that reaction could cause more harm than good for the project. Your decentralized market solution is flawless, it is obviously a viable use case, but the hardware industry and hardware integrators will be taking apart the IoT concept and such publicity will hurt your and the project's reputation.
Well this was just an example. Also I should mention that your client would be locking the funds. And its not dissimilar from "Parking" funds in Nubits, we can offer incentive for parking with manufacturers. One method is to use only one manufacturer at first to simplify the distribution. I'm mostly interested in mesh networks to be honest. That is exactly what I discussed with the other players. My solution works, its not "green addresses" per se but I agree even though it works it may not get adopted. That solution works around bitcoins weakness of trustless doublespend problems. What you are describing is a problem inherent in Bitcoin itself so the only other way is to make the device are party to the transaction in some way so it can trust the RAW transaction is valid. This way it doesnt have to wait for confirmations since it knows its holding a signature, hash or something that would prevent you from spending otherwise. Time locks are not permanent, they are only needed to create a connection with the hardware manufacturers server to prove you arent double spending on one of their devices.
So, no offense taken. I have yet to see a better solution. Perhaps I can come up with an additional one later and PM you when I do. After all knowledge should be free. So we are taking baby steps. Just start with what works and let the industry evolve. That solution works for your particular use case. Users can also do a double deposit contract with device manufacturers in advance. Then they would not need to use my above proposed method.
Bitcoin itself is highly flawed anyways, it doesnt have a very advanced scripting system, there is malleability which always needs to be worked around by protocols, there is major problems with botnets and sybil attacks, there is a scalabiity issue that is very complex to confront (im speaking of bandwidth issues in securing the blockchain) and so on.
With that said, they liked my ideas with mesh networks and added it as IoT which is a misnomer. The mesh netowork idea I had was working with hardware but not necessarily in real time. A good example is Blackcoins COLD Staking. This is something rat4 added some RPC calls for. Now I can do multisignature staking in Halo. We could then, use a device with the staker to sign the second sig. So this is a case where you dont need instant confirmations.
To me, mesh networks are a dream goal if we even get close to that point this project will be a smash hit. So really my impression in starting this project was to start in mesh. If they really want to do this with me, we can change the world so its of great interest to me to see this follow through.
Lets be clear about this roadmap again:
*DEMO Smart contracting client(from BlackHalo) --- Within a week. Trying to get it before ICO ends but not 100% sure.
*Markets(beta without advanced whitelists but with a kill switch/mod key) --- Within a month or two max. I've been working on this for Blackcoin, they coincide
The last two get worked on simultaneously with Hedging taking precedence:
*Hedging transition with the checklocktimeverify --- this is about a month... we can pull from NuBits, but we are dealing with a larger supply so that math has to work out to manage all of everyones funds properly. If they pay my dev team I can have them work on this fork BEFORE I finish my markets which would greatly increase the speed in the fork.
*Markets with whitelists/double back on server --- one month after the main markets release. Its not a difficult implementation, its very elegant but there would also be some user interface feedback and bugs we want to respond to. This could delay one additional month so we will have to ask users to behave in our decentralized markets so we don't have to shut it down.
So give the above a total of 3 months. Realize I'm trying to underpromise and overdeliver. And you know I deliver if you read blackcoins subreddit
ONCE we are hedged, we can go for the mesh networks. We will now be capped at a billion dollars(thats my target price anyways) and be able to afford and utilize our connections to the hardware markets and we will indeed create an incentive system for setting up these networks.