Mozilla argues that Apple's EU terms — requiring a separate EU-only binary, 30-day travel invalidation, no code reuse with the global iOS build, and a new review/provisioning/distribution channel — make shipping Gecko on iOS commercially untenable. Their conclusion is that the structure of the compliance regime, not any single clause, is the deliberate barrier.
The editorial reads Mozilla's unusually blunt tone as confirmation that the fragmentation requirements (region-locked binaries, no A/B testing, no gradual migration) are structured to produce exactly the outcome observed: zero shipping non-WebKit browsers almost two years into the DMA. The point is that the rational business decision under these terms is to not ship at all.
The 222-point thread is dominated by developers who have stopped expecting the status quo to change and are debating whether the DMA has any real teeth. The prevailing frustration is that the regulation was specific — it named engines, it named Apple — and the European Commission still hasn't moved to escalate against Apple's narrow compliance interpretation.
For fifteen years the standard answer to 'iOS Safari is the new IE6' was to wait for regulators. The editorial points out that regulation arrived, was specific, named the right parties, and still produced a rounding error in market response — suggesting the pro-regulation theory of change for browser engine competition has been empirically falsified.
Mozilla's policy team published a post this week titled *Why there are still no non-WebKit browsers on iOS*, and for a company that normally hedges every sentence through three layers of legal review, the tone is unusually blunt. The post walks through, in order, every technical and contractual obstacle Apple has erected between a browser vendor and an iOS user — and concludes that the Digital Markets Act, nearly two years after taking effect, has produced exactly zero shipping non-WebKit browsers on iPhone.
The specifics are worth naming. Under Apple's current EU terms, a browser vendor that wants to ship its own engine must build and maintain a separate iOS binary that only works for users whose Apple ID region is set to an EU country. If a user travels outside the EU for more than 30 days, the engine-capable build stops receiving updates. The vendor cannot reuse any existing iOS codebase — the entitlement to ship a non-WebKit engine is gated behind a new review process, new provisioning, and a new distribution channel. There is no way to A/B test against the WebKit build, no way to migrate users gradually, and no way to serve a single global binary.
Mozilla's conclusion is the quiet part said out loud: the compliance regime is structured such that the rational business decision is to not ship Gecko on iOS at all, which is almost certainly the point. Google has made similar noises about Blink on iOS and has similarly shipped nothing. The Hacker News thread, 222 points and climbing, is dominated by developers who have given up expecting this to change and are now arguing about whether the DMA has teeth or whether the European Commission simply lacks the appetite to escalate.
For fifteen years the standard response to "iOS Safari is the new IE6" was *wait for regulation*. Regulation arrived. The regulation was specific. It named browser engines. It named Apple. And the market response has been a rounding error.
The interesting question is not whether Apple is complying with the letter of the DMA — that's a legal argument that will take years and will be decided by people who have never shipped a Service Worker. The interesting question is what the compliance *pattern* tells you. Apple has demonstrated that a sufficiently motivated platform owner can neutralize a regulatory remedy by making the compliance path economically irrational, without ever refusing to comply. Ship the entitlement. Make the entitlement unusable. Wait out the appeal.
This matters beyond browsers. The same pattern — technically-compliant, practically-prohibitive terms — is now being applied to alternative app stores (Core Technology Fee), to NFC access, and to default-app switching. Browser engines are the canary because they are the most visible and the most expensive case to port. If Mozilla, with twenty years of Gecko investment and a dedicated mobile team, cannot make the math work, nobody downstream can either.
The technical cost of the current regime is real and measurable. Safari on iOS has historically been the slowest tier-1 browser to ship WebGPU, the slowest to implement the full Web Push surface, and the slowest to adopt Container Queries and `:has()` at feature parity with the desktop web. When developers complain about "iOS Safari lag," they are not complaining about Apple's engineers — they are complaining about a monopoly on the pace of the mobile web, and that monopoly is structural, not technical. Every PWA strategy, every offline-first architecture, every WebAssembly-heavy client app has to assume that WebKit's roadmap *is* the mobile web's roadmap on half the devices in the developed world.
The HN thread surfaces a related point that is worth flagging. Several commenters — some ex-Apple, some ex-Mozilla — note that even if the EU terms were rewritten tomorrow, the cost of maintaining a second mobile engine on an iOS-specific toolchain is enormous. Gecko-on-iOS is not a flag you flip. It's a multi-year platform port, a new graphics backend, a new JIT story (Apple still prohibits general-purpose JITs outside WebKit even under the new entitlement, which neuters JavaScript performance from day one), and a new security model. The DMA, as currently drafted, does not force Apple to allow JIT. Without JIT, a shipped Gecko on iOS would be slower than Safari on every benchmark that matters. A regulation that forces a port but prohibits the optimizations that make the port competitive is not a remedy; it is a trap.
If you ship a web app in 2026, your iOS runtime target is still WebKit. Plan accordingly. Three concrete implications:
First, do not design around browser-choice as a future out. If your product's performance ceiling is set by Safari's JavaScriptCore, that ceiling is permanent for the planning horizon of anything you are building this year or next. The architectural decisions this forces — server-side rendering over client-heavy SPAs, native clients for compute-intensive features, careful WASM budgets — are not going away. If you have been quietly betting that "Chrome on iOS will fix it," stop.
Second, feature-detect aggressively and assume Safari is the floor. Container Queries shipped in Safari 16; `:has()` shipped in 16.4; WebGPU is still gated behind a flag in Safari 17.4 and only partially there in 18. If your product targets iOS users at all, your feature baseline is whatever the oldest supported Safari supports, full stop. Progressive enhancement is not a nice-to-have; it is the only strategy that survives contact with the actual mobile web.
Third, if you work on a browser, a game engine, or a WASM runtime targeting mobile, the Core Technology Fee and the EU-only binary requirement are now the real cost model. The Mozilla post is, among other things, a clear signal to any VC-funded browser startup (Arc, Brave, Ladybird, The Browser Company's next thing) that iOS is not a near-term market. If you are planning a mobile browser story, budget for Android-only for the next three to five years and treat iOS as a locked platform until the regulatory landscape materially shifts.
The European Commission has the authority to designate Apple's terms as non-compliant and impose fines of up to 10% of global revenue, escalating to 20% for repeat offenses. The authority exists; the political will is the open question. Mozilla publishing this post — directly, by name, without the usual diplomatic cover — reads as a lobbying document aimed squarely at Brussels. Expect Google and Opera to follow with similar statements before the Commission's next DMA compliance review. Whether that pressure translates into terms that make a non-WebKit iOS build economically rational, or just produces another round of technically-compliant obstruction, is the only thing that matters for the mobile web for the rest of the decade.
Why is Apple so hell bent on preventing alternate browser engines in iOS? Is there a financial benefit for Apple in forcing their own engine on every user? At the surface level this just seems silly so I must be missing something.
There are non-WebKit browsers on iOS but they don't qualify for notarization:- Reynard: Gecko-based, iOS 13+, extension support https://github.com/minh-ton/reynard-browser- Blinker-Fluid: Chromium-based, iOS 12+, content blocker https://github.com/Nodesclock&#
> One of the conditions requires us not to disclose non-public information related to the projectThis is insane. Basically if you have a cool project you want to show progress and build hype around, you're not allowed to. Apple knows exactly how deadly this is
If anti-monopoly regulators were serious, they would require OS vendors to provide a toggle for "let me install software from anywhere I damn well please with zero interference".The fact that they have not suggests that these regulations are performative.
Top 10 dev stories every morning at 8am UTC. AI-curated. Retro terminal HTML email.
You also cannot work on this outside of Europe. Developers have to use VPN or special networks to make iOS think they are in Europe. Otherwise you cannot build and run with a custom engine.It is completely bonkers.Apple is smart enough to just allow this world wide AND still make huge profits.