Skip to end of metadata
Go to start of metadata

You are viewing an old version of this page. View the current version.

Compare with Current View Page History

« Previous Version 2 Current »

Overview

The Analytics Middle Tier (AMT) will truly come into its own as it grows to incorporate more community-driven use cases. Optimally, those use cases will be implemented and shared by community members. The following sections will help you get started on sharing your SQL code for the AMT.

If you have an idea, but do not have the means for developing it, we still want to hear from you: please create a Tracker ticket as described in step 1 of the Contribution Process below.

Planning Your Contribution

Please review the following documentation to learn how to make a high-quality contribution of new views or sample scripts for the AMT:

The Slack #dev-analytics-middle-tier channel is a great place to discuss new ideas before committing them to Tracker.

Contribution Process

  1. Fork the Analytics Middle Tier repository in GitHub. You will be pushing your changes to your fork, and then making a pull request into the official repository.
  2. Create a ticket in the BIA project.
    1. Bugs: please be sure document the problem, how to recreate the problem, expected response, actual response, and the version.
    2. New view: create as a User Story. Please describe the use case in as much detail as possible so that we can understand what to expect and how to help you integrate your view(s) either into the Core Collection or a use case-specific collection.
  3. The Ed-Fi development team will respond to the ticket in Tracker.
    1. If you're creating a new use case collection, the Ed-FI team will suggest a short code name for that collection (e.g., "ews" for the Early Warning System collection).
  4. Develop your code.
    1. Be sure to follow the AMT Standards and Guidelines to the best of your ability. This will ensure a smooth acceptance process.
    2. When ready, follow the GitHub pull request process described in Code Contribution Guidelines.
  5. Once accepted, the Ed-Fi development team will incorporate your pull request into the code base as its earliest opportunity.
  6. As a community contributor, you will not have access to write documentation directly in our Tech Docs site. The Ed-Fi development team will work with you to translate the Tracker ticket into new documentation under the AMT Collections section of Tech Docs.
  7. After your contribution is merged, we will include it in the next release of the AMT.
    1. Because the AMT uses a strict definition of Semantic Versioning 2.0.0, we can produce new releases frequently and end-users will know by the major version number if there are any breaking changes to be concerned about.

  • No labels