[Dev] ShapeShift Implimentation

Is it possible using API data, to literally implement ShapeShift.io into the Qt…

I don’t need to explain the advantages here.

https://www.youtube.com/watch?v=OezDK2O5Dnc

Welcome to ShapeShift…

ShapeShift is a new piece of infrastructure in Bitcoinland. It is how digital currency exchange should work. From start to finish you can change currencies in under ten seconds, no account required.

API Documentation

ShapeShift is primarily intended to be used as an API.

For any of the api calls if there is a problem the error message will be returned as the following JSON:
{error:"Error message"}
Many of the requests require a ‘coin pair’. A coin pair is of the format deposit_withdrawal. Example: ‘btc_ltc’. Valid pairs are any combination of the below listed valid coins.* The list will grow as we add more:

btc, ltc, ppc, drk, doge

* If a particular coin goes offline any pairs using it will return a message stating that pair is temporarily unavailable.

All requests are available via http or https. We recommend using https. http may be phased out in the future.

wrapper

The disadvantages I see are,

1. Not decentralised.

2. Possible point of weakness.

3. It is just an exchange, and charges fees.

4. Don’t know where or who runs it.


Could be wrong tho'

MrWyrm

It goes back to the plug in thing. If it’s a plug in that people find useful, yay. If we build it into the wallet and it all goes wrong and people lose funds, it’s not yay.

Plugin ahoy?

MrWyrm

Arr matey. Hoist the main sail.

lizhi

I think it is great job ,how to start ?

I think it is great job ,how to start ?

Well. The core you are working on. Would it be best to wait untill that is ready and build it on that?

To be honest I’m not sure how this would work.

We would prefer it to be modular. Essentially, it needs to be optional.

If we can build the under laying structure for api based modules into the new core (if that’s how this works) we can make this the first module to be developed.

I also think this raises the question of having the option of self updating wallets and self updating modules. Or at least the ability to push the updates from within the Qt that way there’s need to manually update wallets etc. Can this be achieved at the same time. Is there a way that software could self update via github?

I think that is where we should start.

The disadvantages I see are,

1. Not decentralised.

2. Possible point of weakness.

3. It is just an exchange, and charges fees.

4. Don’t know where or who runs it.
Could be wrong tho'

In time, there will be a plugin to replace this. Once we figure out a decentralised option, it will be good. But this is to keep us in the game so to speak in the mean time.

lizhi

Yes, I will try to support their API in new core wallet. I need about one week to finish the job.

http://cointelegraph.com/news/112912/shapeshift-lens-unveils-automated-btc-payments-using-altcoins

ShapeShift Lens allows Bitcoin payments to be sent using a range of altcoins, with the conversion handled automatically once payment has been submitted.

The free tool is currently available on Google Chrome Webstore, and features a plugin with which users may access the convertor. Once installed, the ShapeShift icon appears next to Bitcoin addresses on any webpage visited with the browser. Clicking activates the Lens, which takes altcoin payments, converts them to Bitcoin and sends them to the desired address.

There you go… with shapeshift, you can buy any btc item online with ftc.

People wanted ftc to be money? Well, here’s our chance.

Wellenreiter

…

To be honest I’m not sure how this would work.

We would prefer it to be modular. Essentially, it needs to be optional.

If we can build the under laying structure for api based modules into the new core (if that’s how this works) we can make this the first module to be developed.

I also think this raises the question of having the option of self updating wallets and self updating modules. Or at least the ability to push the updates from within the Qt that way there’s need to manually update wallets etc. Can this be achieved at the same time. Is there a way that software could self update via github?

I think that is where we should start.

How could it work?

we could compile the modules into libraries and place them into a known directory.

At startup the clients scans for libraries and adds these or better the contained modules to the GUI. A pre-defined function common in all libraries could return the menu-items and other stuff needed.

Self updating wallets

For Linux pre-compiled binaries I would leave that to the operating system update function. Most Linux distributions have an auto-update function, where the repositories of installed programs are scanned for updates regulary and either automated or user triggered updates are possible. New or changed dependencies - e.g. change db4 -> db5 in the 0.8.x core - are solved automatically

Auto-update from Github is close to impossible, as only the source code is available and the user needs to compile the binaries locally

kris_davison

Unless we stored the compiled binaries in a separate github?

Wellenreiter

Unless we stored the compiled binaries in a separate github?

still would not solve the dependencies problem

kris_davison

True

https://www.youtube.com/watch?v=4z7U4kU4NJA

How could it work?

we could compile the modules into libraries and place them into a known directory.

At startup the clients scans for libraries and adds these or better the contained modules to the GUI. A pre-defined function common in all libraries could return the menu-items and other stuff needed.

Self updating wallets

For Linux pre-compiled binaries I would leave that to the operating system update function. Most Linux distributions have an auto-update function, where the repositories of installed programs are scanned for updates regulary and either automated or user triggered updates are possible. New or changed dependencies - e.g. change db4 -> db5 in the 0.8.x core - are solved automatically

Auto-update from Github is close to impossible, as only the source code is available and the user needs to compile the binaries locally

Ok then. Sounds like somewhere to start. Shame about the auto updating thing. Maybe best to not worry about that at this given point in time.

But yeah, In regards to the modules, that sounds like a great starting point. There’s a lot to brainstorm first but this will work out.

lizhi

Today, Shapeshift.io invited us to test. include FTC_BTC and FTC_LTC

Who can test it ?

Hopefully we will have one of their devs on here soon so they will be able to discuss with us the potentials for SS and also Lens.

Once we can have a solid discussion we will see exactly how we shall go about this, but this is now a current project.

Multisig is well under way, stealth address implementation will begin in about a month or so which at the same time, 2FA should be in development also.

lizhi

I’m working for the api of shapeshift . But Now I find that their services are not stable.May be busy or error or expired.

kris_davison

It looks like there is a timeout of about 15 seconds or so for the exchange rate. I think there is something on the API where you can check how long is left until the end of this timeframe.