[DEPR]: Built-in support for third-party services in the frontend

[arbrandes]

RFC Start Date

2025-08-01

Target Plan Accepted Date

2025-08-15

Target Transition Unblocked Date

October 2025

Earliest Breaking Changes Unblocked Date

January 2026

Earliest Open edX Named Release with Breaking Changes

Verawood - 2026-04

Rationale

This DEPR aims to remove explicit, built-in support in frontend-* repositories for optional third-party services, including but not limited to:

• Algolia
• Google Analytics
• Hotjar
• Optimizely
• Segment
• Zendesk
• robots.txt

It should not fall onto our maintainers to support these types of services, for the following reasons:

1. These third-party services more often than not do not align with the goals of the Open edX project;
2. It is not in the Open edX project's purview to pick one vendor over another;
3. It is difficult for maintainers to test that they still work after making changes to the code;
4. Vendors will on occasion introduce breaking changes to their API (see Google Analytics v3 -> v4), forcing maintainers to effect upgrades where resources could be better allocated elsewhere;

Instead, these should be implemented (and, optionally, published) as frontend plugins by the organizations that rely on them.

Removal

The first set of removals will happen as part of the frontend-base migration. What follows is a list of per-repository issues, each of which list what is being removed from that particular repository:

• Convert Authn to frontend-base frontend-app-authn#1498
• Convert Learner Dashboard to frontend-base frontend-app-learner-dashboard#657

This list will be expanded as the frontend-base migration progresses.

(The issues include more than what this DEPR covers; other DEPRs will be filed for the difference.)

Replacement

There will be no direct replacement for any of the removed code. Open edX users that have a continuingneed for one of the removed services are encouraged to write (and publish!) frontend-base-styled plugins for them.

If changes to the frontend extensibility architecture of the project are required for any such plugin to work (such as adding slots to frontend-base, or creating a tracking abstration layer) maintainers will happily review PRs that introduce them.

Deprecation

As noted above, these removals are already being made in long-lived frontend-base feature branches, one per repository. These branches are intended to either be merged to or become master in time for the Verawood release, which will be cut in March or April 2026, but not earlier than January 1st of that year.

Any contributions meant to extend the extensibility architecture to support plugins that implement functionality being removed as a consequence of this DEPR should ideally target the aforementioned feature branches. At the very least, though, they should aim for a merge before the cut of Verawood.

Additional Info

The slot API that frontend-base exposes is not compatible with the one that exists today. Plugin authors that intend to write replacements for the removed services should familiarize themselves with it.

Task List

No response

Task List

No response