Rebuilding Fulldive VR: Unity Migration and Legacy Repairs

Written by the Fulldive product engineering team, based on direct inspection of the repositories listed below.

Fulldive VR history is not a straight line from launch to present. Looking at mobile VR app evolution through the fulldiveVR/fulldive-unity repository, plus the supporting fulldiveVR/MediaVR and fulldiveVR/fulldive-unity-plugins repositories, the 2022-2024 period is dominated by one activity that rarely gets a marketing bullet: repair. Feeds that had broken were restored, build pipelines that had stopped producing green artifacts were fixed, portal visuals were imported back into the scene, and in 2024 the whole app moved to a newer Unity version. This post reconstructs that rebuild from commit evidence.

The repositories involved

Three repositories carry this story:

fulldiveVR/fulldive-unity — the main Fulldive VR Unity app (package in.fulldive.shell), repository history from April 2017 through October 2024 across roughly 2,220 commits.
fulldiveVR/MediaVR — a much smaller repository, February 2018, four commits, framed as sample/module code.
fulldiveVR/fulldive-unity-plugins — a small Android build/plugin support repository, October-November 2024, five commits.

The main app drives the product story; the other two show that the engineering work extended beyond the one Unity project.

2022: restoring legacy feed and dive requests

The restoration pattern in 2022 is explicit in the commits:

963bc03e restored the old feed request.
ba6dc4a6 restored and fixed the old dive request build.
90d360d0 imported the portal VFX and screen logic back into the app.

“Restored” is the telling word. Feed and dive request code had either been removed, broken by a refactor, or diverged from a back-end contract. Bringing it back was a deliberate choice to keep the 2020-era social and discovery features working on top of a scene that had continued to evolve. Portal VFX — the visual transition when a user moves between VR spaces — is the kind of thing that silently degrades during a long refactor; the import commit shows it was deliberately reintroduced rather than left broken.

This matters because it tells us the product team chose continuity with the historical Fulldive VR experience rather than shipping a simplified new shell. The 2020 reactions, comments, and bookmark features described in VR media discovery remained usable because the feed and dive request paths were repaired.

2023: the less glamorous build-fix year

2023 in fulldive-unity is a build-pipeline year. Commits 01f20dc8 and 6f220ec5 are build process fixes. A Unity project that was authored against one version of Android tools, one Gradle, one JDK, and one set of plugin compatibilities produces surprising failures when any of those move — and they moved a lot in 2023 as Google Play tightened target-SDK requirements.

This kind of work produces no visible user-facing change. It shows up as: the next store build could actually ship. That is the only metric it needs to hit.

2024: the Unity migration

The 2024 highlight is dfc07c05: a migration to a newer Unity version. For a project with a seven-year history and more than two thousand commits, an engine upgrade is an intrusive event. Behaviors change across scripting runtime, rendering pipeline defaults, package manager versions, and Android build tools. Asset import paths can shift. Shaders can need adjustment. The fact that the migration commit lands alongside continued product work is the important signal: the team chose to invest in the engine rather than let the app drift.

In the same late-2024 window, fulldiveVR/fulldive-unity-plugins appears as an explicit support repository. It starts on 2024-10-15 (281325c), shows a 6.6.0 version marker on 2024-10-24 (b89c357), goes through build-artifact cleanup on 2024-10-28 to 2024-10-31 (db3d0f4, f47a287), and lands a 6.8.0 marker on 2024-11-02 (f3b434d). Those versions are repository-internal; we are not claiming they match any specific store version without checking the current listing.

The plugin-repo existence itself is the evidence that matters: Unity migration work pushed Android build and plugin wiring into its own maintained repository rather than living inside the main project.

Where MediaVR fits

fulldiveVR/MediaVR is the smallest of the three. Its full recorded history is February 6-9, 2018: initial commit and README (53c5715), README update (2c021e0), copy from a working branch (0ddeff7), and a sample-app framing (19c56f7). It is framed as a sample app demonstrating Fulldive VR code usage.

We are including it here because it explains something that would otherwise be confusing in the repo map. MediaVR (com.fulldive.mediavr) is not the source of the Play Store Fulldive VR app; in.fulldive.shell is. MediaVR is supporting context: evidence that the team experimented with reusable VR media modules alongside the main product. Treat it as engineering background, not as an app you can install.

What the rebuild says about mobile VR app evolution

Three patterns across the repositories:

Continuity bias. Features built during the 2020 social-layer cycle were repaired rather than removed. The team chose to keep Fulldive VR recognizably the same product through the rebuild.
Build investment. 2023’s un-flashy build-fix commits and the 2024 plugin repository together are why 2024’s Unity migration was possible at all.
Scope honesty. MediaVR did not become the main app. It stayed a sample. That is the right decision to make when a module is not production-grade, and repo evidence shows the team made it.

What this does not do

This post is repository archaeology. It does not describe current store-listing behavior, current headset compatibility, or a specific commercial roadmap. Before installing any Fulldive VR build, users should check the current Fulldive VR project page and the store listing linked from it. Unity and Android platform requirements continue to change, and an engine or SDK migration is a routine activity for any maintained mobile VR app, not a one-time event.

How Fulldive VR plays 360 and 3D video on Android covers the player pipeline preserved across the rebuild.
Why mobile VR apps need headset-specific work explains why engine migrations are coupled to SDK updates.
VR media discovery: reactions, comments, and bookmarks describes the features that were restored in 2022.
What a mobile VR app can and cannot do in 2026 sets honest expectations for today.
The Fulldive company story traces how the VR work seeded the rest of the app ecosystem.

Sources

Repository evidence: fulldiveVR/fulldive-unity, commits 963bc03e, ba6dc4a6, 90d360d0, 01f20dc8, 6f220ec5, dfc07c05.
Repository evidence: fulldiveVR/MediaVR, commits 53c5715, 2c021e0, 0ddeff7, 19c56f7.
Repository evidence: fulldiveVR/fulldive-unity-plugins, commits 281325c, b89c357, db3d0f4, f47a287, f3b434d.
External reference: Unity Upgrade Guides for engine migration context.
External reference: Android Developers — target API levels for 2023 Play target-SDK pressure.

Last updated: 2026-04-16. Commit hashes and version numbers are drawn from Fulldive repositories inspected on 2026-04-13.