Skip to content

Latest commit

 

History

History
127 lines (84 loc) · 6.06 KB

README.md

File metadata and controls

127 lines (84 loc) · 6.06 KB

Overview

NuGet version (chorus) Build, Test and Pack

Chorus is a version control system designed to enable workflows appropriate for typical language development teams who are geographically distributed. These teams need to edit a set of common files, working towards a publication. They want to share their work while, crucially, being able to defer dealing with conflicting edits for periods of time or until a qualified team member can make decisions about the conflicts. The system is implemented on top of a commonly-used Open Source Distributed Version Control System. It works in scenarios in which users are connected by Local Area Network, Internet, or hand-carried storage devices. Chorus supports several workflow models, including those that maintain a “master” database separate from the incoming submissions of team members. Quite unlike the version control systems commonly in use, Chorus works invisibly for the common cases and is kept simple for even beginner computer users.

Distinctive Features

These features come for free with any Distribute Version Control System:

  • Share files between users, even if they are never connected to the internet.

  • Every member of the team has access to a full history of all work done by the rest of the team.

  • In a crisis, work can be "rolled back" to a previous version.

However, "raw" Distributed Version Controls Systems are relatively difficult to understand, configure, and use, even for computer-savvy workers.

The following list of features should help you understand why we built this layer over a raw version control system:

  • silently synchronize; will never tell the user to manually merge conflicts first

  • automatically check for team members & devices with which to synchronize

  • Support a Master branch which does not automatically accept changes from anyone

  • Files can be marked as shared by the team or user-specific. This allows things like preferences/configurations to be part of the repository but kept separate for each individual. This will also allow one team member to make configuration changes for another, remote member, and push those changes through the system to that user, without physically accessing their computer.

  • 3-Way, schema-savvy XML merging. Various policies can be implemented for choosing a winner in the case of conflicts. Regardless of the policy, details of the conflict are logged in an XML file which also under version control. At a time and place of the team's choosing, these automatic choices can be reviewed and reversed.

  • Configuration help from applications. Applications generally know where their important files are, which files are individual-specific, and which should not be backed-up/shared at all. Applications that know about Chorus pass this information to it, so that users don't need to become experts in how all the files work.

  • Synchronization help from application. Applications often know what points are good ones for checking data in. For example, when exiting, or before doing a large and possibly undesirable operation, like deleting a large number of items or importing a new data set.

  • In-Application conflict and change history. Rather than ask users to learn version-control-specific tools, the Chorus model is that Chorus provides the raw information applications need to provide a smooth, integrated workflow in the same environment as the user has for editing. For example, a dictionary-editing program using Chorus will allow the user to see a full history of the current record, including who made what changes, and what conflicts (if any) were encountered during synchronization.

  • A built-in "notes" system which makes it very cheap to give users the ability to add notes to any piece of data, and to carry on conversations about the data until they mark the note as "resolved".

Status

Chorus is functional and being used in several applications with different development teams. However, we are not really interested in supporting any further uses until things mature and someone writes good developer documentation. Documentation (what little exists) drips out in the form of occasional blogs here.

Testers

Please see Tips for Testing Palaso Software

To send and receive with the test server over the Internet, set the following environment variable:

LANGUAGEFORGESERVER = -qa.languageforge.org

To register to use the test server, visit https://public-qa.languageforge.org

Developers

Road Map & Workflow

https://github.com/sillsdev/chorus/issues

Coding Standards

Palaso Coding Standards

Source Code

Chorus is written in C#. The UI widgets use Windows Forms, but you could make your own using a different platform and just use the engine.

After cloning the project you should now have a solution that you can build using any edition of Visual Studio 2022, including the free Express version, JetBrains Rider or Visual Studio Code. This works on both Windows and Linux.

From the command line you can build with:

dotnet build

and run the unit tests with:

dotnet test

Building client projects against locally-built artifacts

  • Set an enviroment variable LOCAL_NUGET_REPO with the path to a folder on your computer (or local network) to publish locally-built packages
  • See these instructions to enable local package sources
  • build /t:pack will pack nuget packages and publish them to LOCAL_NUGET_REPO

Further instructions at https://github.com/sillsdev/libpalaso/wiki/Developing-with-locally-modified-nuget-packages