augur | client

Augur is a decentralized prediction market platform built on Ethereum.
It runs locally in your browser and communicates directly with the ethereum network, without going through intermediate servers.

There are several ways to run it:

• Easiest: visit dev.augur.net
• Best: run ethereum locally

Requirements

Git
Node or Docker

A Note to Windows 10 Users:
Turn on Developer Mode and also enable Windows Subsystem For Linux so that you have access to bash.
Run all subsequent commands within the bash command prompt.
It's also recommend that you use a Debian based workflow for the installation of packages (makes things much easier).

Build From Source

git clone https://github.com/AugurProject/augur.git
cd augur

NPM

npm i
npm run build

Yarn

yarn
yarn build

Docker

docker build -t augur .

This will create a build folder inside of the augur directory with all the files necessary to run the client. Simply copy these files to your web server of choice.

Develop

NPM

npm i
npm run dev

Yarn

yarn
yarn dev

Docker

To run augur ui in docker map your external port (8080) to docker container port 80

docker run -p 8080:80 augur

Visit http://localhost:8080

Run Ethereum Locally

There are a number of ways to go about running an Ethereum local node against Augur, described below is the CLI method using geth.

Start off by installing geth.

Depending on how you'll be running Augur (Development OR Production), follow the corresponding steps below:

NOTE -- Unless you have a specific reason for doing so, generally you should not simultaneously run geth and another Ethereum client (such as Mist or Parity).

See more detail notes to run augur UI on local ethereum node and augur-node dev-local-node

Development -- Testnet (Rinkeby)

Create an Ethereum account (if needed):

geth --rinkeby account new

Be sure to securely store your password and remember the account address (displayed after account creation).

Start geth:

geth --rinkeby --unlock 0 --rpc --ws --rpcapi eth,net,shh,admin,txpool,web3,personal --wsapi eth,net,shh,web3,admin,txpool,personal --rpccorsdomain '*' --wsorigins '*' --cache 2048 console

Additional points of note:

• The value supplied to --unlock should correspond to the account you'd like to be using for Augur transactions.
  • To get a list of accounts and their numerical value run geth account list.

Production -- Main Network

Important note: the Augur contracts have not yet been uploaded to the live Ethereum network, and therefore Augur is not yet available on the mainnet. The following instructions are here as a placeholder for when our contracts are uploaded to the mainnet.

Create an Ethereum account (if needed):

geth --rinkeby account new

Be sure to securely store your password and remember the account address (displayed after account creation).

Start geth:
Important -- The following command will be using real Eth (Main Network) and is potentially more permissive than necessary for your use case.
DO educate yourself surrounding the arguments and determine which are appropriate for your specific use case.

geth --unlock 0 --rpc --ws --rpcapi eth,net,shh,admin,txpool,web3,personal --wsapi eth,net,shh,web3,admin,txpool,personal --rpccorsdomain '<domain of server>' --wsorigins '<domain of server>' --cache 2048 console

Testing local geth node

curl --data '{"method":"net_version","params":[],"id":67,"jsonrpc":"2.0"}' -H "Content-Type: application/json" -X POST localhost:8545

returns network id: should be 4 for rinkeby
{"jsonrpc":"2.0","id":67,"result":"4"}

Additional resources:
JSON RPC API Documentation
Javascript Console

Additional points of note:

• The value supplied to --unlock should correspond to the account you'd like to be using for Augur transactions.
  • To get a list of accounts and their numerical value run geth account list.
• The values supplied to --rpccorsdomain and --wsorigins should correspond to the server's domain which is running the instance of Augur. BE SPECIFIC + DO NOT USE A WILDCARD

Documentation

http://docs.augur.net

Development Tips

#1: Search the project filenames for whatever you are looking for before thinking about it too much.

Since our code is structured in many small files that are named the same as the state/functionality they represent, rather than try to follow and reverse engineer code paths, just blindly search all filenames for whatever it is you are looking for. More often than not, you will find what you need.

Want to know where the css for pagination is? Don't bother tracing where/how they're included, just search your files for pag >>>and pagination.less will pop up.

Want to see how the login account gets updated? Search the files for login >>> and update-login-account.js will appear.

Sponsorships

As an open source project, we'd like to thank BrowserStack for providing free access to their environment testing platform!