Inside the OSphreX feasibility study

The OSphreX feasibility study is an 86-page document. This is the short version.

The question

If a small team — one to three engineers, no venture funding — sets out to build a privacy-first mobile OS from a clean architectural slate (Linux kernel, Wayland compositor, custom UI shell, not Android), what’s the honest timeline to a tech-preview daily-driver?

Not Android with extra privacy. Not a custom AOSP fork. A different OS, structurally.

The findings

No mobile OS in the last 15 years has been built truly “from scratch” by a small team and reached daily-driver status. Every survivor — Sailfish, Ubuntu Touch, Firefox OS, KaiOS, webOS, postmarketOS, Mobian, Droidian — reuses the Linux kernel and large parts of the freedesktop middleware (Mesa, PipeWire, NetworkManager, ModemManager, systemd). Only SerenityOS attempts true from-scratch OS development, and it has taken six years on x86 desktop and isn’t yet a daily-driver.

The Honor Magic 4 Pro is effectively a closed device for OS bring-up purposes. Honor stopped issuing bootloader unlocks to consumers in 2022. Only “test devices” sold to specific channels can be unlocked. We picked the Magic 4 Pro as a design target — an aspirationally-spec’d flagship to size the project against — but actual bring-up has to happen elsewhere.

The recommended bring-up pivot is Xiaomi 12 (cupid) or Nothing Phone (2). Both use the Snapdragon 8 Gen 1 (SM8450). Both have unlockable bootloaders. Both have active mainline Linux kernel and Mesa work via the postmarketOS community.

The honest small-team roadmap is 30–36 months to tech-preview daily-driver. Year 1 reaches Wayland boot, screen, Wi-Fi, USB charging. Year 2 reaches SMS, voice calls, notifications. Year 3 reaches camera, polish, app store, OTA, “Scripts tab v1” (the customisation primitive at the heart of OSphreX). This matches the postmarketOS / Mobian historical curve — it is not aspirational. The original PinePhone took the community ~5 years to reach “usable but flaky.”

What we’d pivot on

If by month 18 we haven’t reached SMS / voice / notifications on the chosen bring-up device, the project becomes archive plus design study rather than path to product. The feasibility study commits to that decision criterion in advance.

What this is for

OSphreX is not Apple’s next chapter of iOS, and it is not Google’s next chapter of Android. It is a research initiative asking what an alternative would look like if a small team built one carefully. The feasibility study is published in full so other small teams can read the work and either join, fork, or argue.

Read the OSphreX research initiative for the longer summary, or download the full feasibility report (linked from that page).