https://github.com/starptech/judge-framework

Judge-Framework for decision making in large teams - http://judge-framework.starptech.de/
https://github.com/starptech/judge-framework

Last synced: 21 days ago
JSON representation

Judge-Framework for decision making in large teams - http://judge-framework.starptech.de/

• Host: GitHub
• URL: https://github.com/starptech/judge-framework
• Owner: StarpTech
• License: mit
• Created: 2020-02-24T11:08:10.000Z (over 4 years ago)
• Default Branch: master
• Last Pushed: 2023-01-07T15:12:37.000Z (almost 2 years ago)
• Last Synced: 2024-10-03T23:43:00.104Z (about 1 month ago)
• Language: JavaScript
• Homepage:
• Size: 2.3 MB
• Stars: 9
• Watchers: 3
• Forks: 0
• Open Issues: 14
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# judge-framework

Judge-Framework for decision making

The framework was shown first in [code.talks commerce 2018 - Agile Transformation benötigt Einhörner](https://www.youtube.com/watch?v=t4eVn_MxOUQ)

## What is it?

You can share a link with your colleagues to get realtime evaluation of the voting.

## How does it work?

We use MQTT protocol to build a pub/sub model. All messages are transfered through the public [hivemq](https://www.hivemq.com/public-mqtt-broker/) MQTT broker.

We don't collect or send any sensitive data.

### In detail:

**Challenges:**

- Share a unique link

- Distribute votings to all clients

- Sync new clients which joins later on

- Protect against double votes

- Protect against duplicate and out of order messages

### Browser state

When a user shares a link we create a new unique subscription in form of `judgeframework1.0/{group-id}` and store the group id in the url.

### Distribute votes

When a user visit the link and clicks on a voting card the `decision` is published to the subscription in (Step 1) and all participants will receive the event in order to be up-to-date.

### Sync new and existing clients

In order to redeliver the latest state of the voting to clients which are out of sync we use the `retain` mechanism in the MQTT protocol. It allows to mark a published message as `retain`. A retain message is automatically published to all new and existing clients. The message contains the complete state of the latest voting. After each vote, the retain message gets overwritten and all following clients will receive the latest voting state.

### Double votes

A user can't vote twice we persist the latest voted session in the local storage.

### Message duplicates

Votes from clients can arrive more than once. We maintain a map of voted clients to protect against that.

### Message order

Message order mustn't be guaranteed we are only interested in messages which increase the current state of a card 0-6.