Burst wallet Lite, (Idea for new wallet)


#1

For all new users there is a tedious task to get going with burst. You either need to download a local wallet to then struggle with java to set it up. After that you need to download the blockchain which can seem to take forever for some people. If you do not use a local wallet you need to use an online wallet. With this there is some security trust issues since the online wallets can be mischievous and steal your passphrases. So to go around both of this issues I think it would be a good idea to make a wallet lite that solves above issues. This means that you just download a wallet. Start it and you are up and running directly.

So how can this be achieved?
Well we need to have the blockchain stored somewhere so my idea is that instead or at the same time as one hosts an online wallet you use a new version of the wallet that has some new api functions. And this will act as a server part of the wallet.

When the new wallet accesses the server part it can request blocks only for a specific address so the user can see and view his transactions. The new wallet also has function to create announcement packages and send those to the server to be renounced out to the network. In this way the wallet never sends any password to anyone and is only used within the wallet. And users can safely commit transactions.

I am not sure this wallet type is ready for burst since I still think we need more nodes in the network and this wallet does not make that happen. But it would for sure ease up a lot for newcomers.

Please give your feedback to this and maybe some can work together to see this in reality.


#2

What would happen if the one server where the passwords are kept goes offline?


#3

@Quibus I think what you’re talking about is that Burst needs a “thin” client…coins such as bitcoin and etherium have this. I do believe that that it is a future goal of the PoC Consortium…The network does need some work before then though.

I completely agree about the benefits of such as wallet. This would open the gates to a much wider audience, and also makes things a lot easier on current users.


#4

The password are still on local wallet. not on the servers. the only thing the servers does is holding the blockchain and relaying precomputed anouncements


#5

Correct. but i did not know the that PoCC was developing one.


#6

What if every ten blocks or so a script ran and copied a known good wallet db, zipped it and then set it aside for a call from a new wallet download. Then when a new user downloads the wallet from github as part of its install it transfers in the up to date db and then there is no waiting for the genesis block sync?


#7

would work as long as you dont force a fork on someone, and you also would get a huge download in the end.


#8

They’re not developing one per se…but it is something they will look at in the future. Like I said, there are probably bigger things to deal with until then :slight_smile: Exciting times though eh? It’s something I have hoped for for awhile…an organized push to develop Burst and bring it to new heights.


#9

Well since we do not know for sure if and when PoCC will make a wallet like this i have started to look into this a bit. so far i am not long into the project but i have managed to implement Curve25519 and the nxt implementation of Reed Solomon into my project. With this i can make all account operations.

Steps that are left is:
Create a transaction system so i can send burst.
Make a web api that should be run on a server where a working cg wallet is and share its database with my api.This untill or if someone implements api code to wallet.

I am writing this wallet in .net so it will function on windows platforms and probably Linux with mono framework installed.
The project will be open source when I have a wallet that actually can recive and send burst.


#10

As @ryanw has mentioned, this is called a “thin client”. Such a local wallet does not need to have the whole blockchain available on disk to verify network operations. Technically, this happens via SPV, Chapter 8. SPV is supported for well-known cryptocurrencies by providing bloom filters for block headers. Currently the Burst node does not support this. This has to be implemented first, before developing a thin client. Check out https://github.com/ac0v/burstcoin for updates.


#11

Well, Obviously i have another aproach.