Nicholas Foundation
← All research initiatives

Research initiative

Privacy-First Mobile OS

"Can a small team reach a daily-driver alternative to Android in 30–36 months?"

"Your device that you bought, your choice, not corporate choices."

OSphreX is a research initiative asking what a mobile OS would look like if it were designed around the premise that *the device you bought is the device you own* — and built by a small, principled team rather than a billion-dollar consortium. The feasibility study is published in full; the implementation is the open question.

Current findings

  • No mobile OS in the last 15 years has been built truly "from scratch" by a small team and reached daily-driver status on contemporary hardware. Every survivor (Sailfish, Ubuntu Touch, Firefox OS, KaiOS, webOS, postmarketOS, Mobian, Droidian) reuses the Linux kernel and large parts of the freedesktop.org middleware. Only SerenityOS attempts true from-scratch OS development.
  • The Honor Magic 4 Pro is effectively a closed device for bring-up purposes — Honor stopped issuing bootloader unlocks to consumers; only "test devices" are unlockable. The Magic 4 Pro is therefore a *design target*, not an actual bring-up device.
  • The recommended pivot is Xiaomi 12 (cupid) or Nothing Phone (2) — both have unlockable bootloaders and active sm8450-mainline / mainline Linux kernel device-tree work.
  • The honest small-team roadmap is 30–36 months to a tech-preview daily-driver — matching the postmarketOS/Mobian historical curve. SerenityOS took 6 years on x86 desktop; PinePhone took the community ~5 years to reach "usable but flaky."

What we want to achieve

  • Reach a Wayland-booting demo on a Xiaomi 12 / Nothing Phone (2) within ~12 months — display, Wi-Fi, USB charging, no telephony yet.
  • Reach SMS + voice-call functionality and a notifications daemon within ~24 months.
  • Reach camera (basic libcamera pipeline), polish, battery/thermal tuning, app store, OTA, and Scripts tab v1 within ~36 months — the "tech-preview daily-driver" milestone.
  • Publish every architectural decision openly, including the ones we change our mind about.

What we’re investigating

OSphreX is not a shipping product. It is a research initiative asking what a privacy-first mobile OS would look like if it were engineered from the ground up to honour the premise that the device you bought is the device you own.

The feasibility study (linked above) is the first formal research output of The Nicholas Foundation. It asks four questions, in order:

  1. Is it feasible at all? (Yes, conditionally — see findings.)
  2. What hardware should we target? (Honor Magic 4 Pro as design target; Xiaomi 12 / Nothing Phone 2 as actual bring-up.)
  3. What software stack? (Linux kernel + Wayland compositor — not Android. Custom Rust + GTK4 / libadwaita UI shell.)
  4. What timeline? (30–36 months to tech-preview daily-driver — matching the postmarketOS curve.)

Why this matters

Three reasons, again in increasing order of importance.

  1. Bootloader politics. Google and Apple have made device ownership progressively more conditional. The Honor Magic 4 Pro story (consumer bootloader unlock revoked mid-product-life) is not unusual; it is the trajectory of the entire industry.
  2. Privacy as architecture, not feature. GrapheneOS demonstrates that even an AOSP-based privacy-respecting OS requires monumental ongoing engineering — and is still downstream of Google’s decisions. A clean-slate alternative isn’t easy, but it is structurally different.
  3. Small-team viability. If we cannot honestly say that a small principled team can ship a daily-driver mobile OS in three years, then we should be honest about that — and OSphreX becomes an archive rather than a roadmap. The feasibility study says: it is hard, but it is doable.

What the feasibility report concludes

In short: yes, it’s doable in 30–36 months by a small team — provided we (a) pivot off the closed Honor Magic 4 Pro bring-up target to a mainline-friendly device, (b) accept that the postmarketOS/Mobian timeline is realistic for a small team rather than aspirational, and (c) publish honestly along the way about what works and what doesn’t.

The full report covers hardware bring-up, kernel mainline status, compositor choice, app ecosystem strategy, camera pipeline pivot, and a 36-month indie-team scope roadmap.

Read about OSphreX →

Backed by