Development Update #114

Summary:

We need to start building out the community.
We need to get the three white papers done and the website:
Things we need:

Skycoin Website:

The website is a flat file. No-script (no javascript).

This was first version of website, which was visually unacceptable.

Second version is much better.

In the third version:

Aether: Skyhash

This is the new internet

The first version of Skyhash/Aether is done:

There are no servers. It is completely peer-to-peer.

This can be used for:

We trying to get a developer to work on the content browser for Aether.

Instead of going 120 hops in the network, to get data. It will just get the data from whoever is near you that has copies. - in the current internet, if a content publishers server is in Los Angeles and the requester is in Hong Kong, ever content request and packet will go from Hong Kong to Los Angeles, for each request. - in Aether, the first copy of the data goes from Hong Kong to Los Angeles, and each subsequent copy can be served from a local content cache or another user

This is also called “Content Addressable Storage” or “Source Independent Networking”.

For instance, a file is requested by the SHA256 hash of the data contained in the file. You do not care where the file is downloaded from, as you can verify the contents of the file once it has been received from any source.

If you are using 100 MB/s over 100 hops of fiber (going through 100 routers). You are using 20x the network resource usage, as if you are going through only 5 hops - resource usage is number of hops times bandwidth usage per hop

So by pushing content to the edge, the aggregate network resource usage is reduced and latency is reduced.

This also enables Multi-cast and enables a higher level of network redundancy (automatic geographically distributed content mirroring).

The next generation HTML, websites, services, social media and video streaming will be built on this principal. The existing model of a CDN is an intermediate stage of the evolution of this type of “EdgeNet”. The military, commercial and home IoT and the next generation of “cloud” services will largely be enabled by networking protocols designed on these principals.

We have our object format and serialization format defined and are starting to integrate it into Aether.

We are building the skycoin consensus and block distribution on top of this model, because of the security and speed advantages. It also has the advantage of being a simple three packet protocol.

Aether: Tox-like messaging

Aether at the basic level, enables cryptographic, peer to peer replication pub-sub channels.

We can take two public keys - Have pubkey A subscribe to pubkey B’s pubsub channel - Have pubkey B subscribe to pubkey A’s pubsub channel

So we now have a two way channel:

The implementation is completely in golang, so no buffer overflows.

The messages are stored, by everyone replicating the public key

This is a very powerful messaging primitive. Especially when you have mobile devices or an unreliable network that is going on and off. This type of network can deliver messages, even over latencies of minutes, hours or months. It is - Source independent - Synchronous - Scale invariance in time

For instance,

The latency can literally be days or months and the protocol still operates reliably (scale invariance in time).

Also, all applications built upon the distributed object store, will operate automatically offline with no additional coding - Google maps works while you have internet - When you lose internet, you no longer have a working map - When a map application is run off of map data in the asynchronous, distributed object store there is no difference to the application between being online and offline. You can cache the map for a whole country locally and it will operate the same regardless of internet connectivity.

I think this will find a lot of applications for drones (drone swarms, mobile communication nodes), satellites, mobile devices (cell phone applications), IoT or mobile networking.

Meshnet

The data channel and control channel is separated now. We have three things left to do, before its working again. It could be one or two weeks, before the full stack is working end-to-end.

I think we will have someone get the VPN application working again, so that we will have at least one useful application to run over the network.

The long term goal of Aether is to create content and website that are internal to the network only.

Aether will look exactly like IPFS

Roadmap:

Our application stack is Skycoin + Skywire (Meshnet) + Aether (Skyhash) + CX (application language)

We almost have the coin completed. Then Skywire and Aether in development/demo state.

No translation bounty

Discuss this post on telegram

Skycoin Telegram