have been very busy indeed. Here are two web services I set up:

https://wiki.anisotropic.systems/
https://gogs.anisotropic.systems/dawn-network

Chat

I did have a 'mattermost' chat set up but we have now moved to riot, this web address gets you there:

https://chat.anisotropic.systems

Dawn, Phase One

The way I see it is that Dawn's first blockchain is going to be @faddat's project 'glogchain'. It takes a different approach to Steem, the blockchain only stores references and the data itself is stored on distributed data stores, there can be comments and of course most of the new Dawn tokens will go to the authors winning the most votes. The payout scheme has not yet been hashed out.

One key objective is that all nodes capable of querying the database will themselves produce the webpage you see. After all, it is mostly node.js libraries loaded from canonical repositories, and our page fills in the details to get you the page.

For the more technically minded, this approach will be taken with all of the parts of the system, meaning simple URLs can produce queries from the database in a format that you can process with another application to produce an interface. By default there will be one that works with a web GUI framework it hotlinks to.

The first dawn currency will be built on the core of the basecoin component of Tendermint. We are digging into the code hard, but I expect a couple of months at least before we have a fully running chain.

My project

I have secured myself some relatively decent accommodations and I am busying myself learning how to code in Go and devising schemes for what I consider to be Phase Two of Dawn. I have designed a token ledger protocol that works both offline and allows even small clients to have high confidence in the accuracy of the ledger, by borrowing the model of real life coin marketplaces.

First thing I am doing is building a core, fallback type network coordination system built on Tor. Tor hidden services are similar cryptographically to bitcoin wallet addresses, and each address requires a node to possess a secret key, the address itself is a public key.

So, it is a network that has some pretty serious use cases these days, and it's not going away, and it's at least as good as the channel you can get through almost any firewall, to port 53, the DNS layer. This protocol is like this (in that it helps you find servers on the network), except it just distributes a database containing all nodes that are operating this protocol, and then on top of that I build a second layer that allows nodes to have other tags that associates them.

The objective is to create a utility protocol that allows the creation of a secure, anonymised global network connectivity system for distributed systems, and that then allows an application to claim an address space that all nodes of a type can find, with only one address from one active node.

This is small scale stuff, because I am very focused on getting these kinds of details right if they are going to become widely used. I should only take a week or two to build the node, and it will become a component of the total system, that I will build up piece by piece.

TL;DR:

Simply, I am building a thing that allows the servers to find each other anywhere on the Tor network, and identify themselves as some particular type of server, part of that system's network.

Simply getting nodes together in distributed systems on the network can be very difficult. Really it is a use case that current business models don't have a lot of use for, but it is one that a lot of new business models can use (sharing economy).

My witness server

Currently, it is earning about twice as much as it costs, so I am leaving the witness running. I set my price feed at roughly hourly intervals from coinmarketcap, and right now it is a black box that I can depend on seeing a current price published and no missed blocks.