EOL #001

Art of FACELESS | epicFAIL#

Apple has announced that Rosetta (the translation layer that allows Intel-based apps to run on Apple silicon Macs) will be killed with macOS 28, due in late 2027. If you missed the notification, that's understandable. Apple buried it in a support document, framed it as helpful advice, and dressed it up in the language of progress. The usual tech guff.

It is not progress. It is the latest iteration of a pattern that has been running for over two decades. And if you are a creative professional who invested serious money in a Mac ecosystem because you were told it was stable, reliable, and built to last, you have been here before.

You will be here again.

The Pattern Has a Name

In product lifecycle management, EOL means End of Life. It is the moment a manufacturer formally stops supporting something you paid for. In pharmaceuticals, in engineering, in any discipline where methodology is validated and documented, EOL is a scheduled, transparent process with clearly stated criteria. You know it's coming. You plan for it. The decision is justified by evidence.

In consumer technology, EOL means something different. It means the decision was made upstream, the rationale is yours to speculate about, and the timeline is whatever suits the manufacturer's product cycle. Your workflow, your investment, your decade of accumulated tools, these are externalities.

Apple has run this play three times in twenty years.

2006: PowerPC to Intel. The transition was announced in 2005. Rosetta 1 was introduced to ease it. By 2011 it was gone, taking with it a library of creative software that had never been ported. If you were running legacy audio, video, or print tools, you either paid to upgrade or you stayed on an ageing OS indefinitely.

2019: macOS Catalina kills 32-bit. Apple had been warning developers since 2018 that Mojave would be the last OS to support 32-bit applications. When Catalina dropped, at least 235 apps stopped working overnight. Not slowed down. Not warned. Stopped. Adobe CS5, legacy audio plugins, independent software tools built by small developers who could not afford the cost of a full rewrite, gone. The official advice was to contact the developer. Many developers no longer existed.

2026: Rosetta 2 entering end of life. The first M-series Macs appeared in late 2020. Rosetta 2 was introduced to ease the Intel-to-Apple-silicon transition. In 2025, Apple confirmed it would remain available through macOS 27, then be restricted in macOS 28 to "certain older, unmaintained games." Every other Intel-based app: unsupported. Update or replace. The clock is running.

Three architectural transitions. Three compatibility layers introduced. Three compatibility layers terminated. Three software graveyards.

The Hardware You Cannot Fix

The transition argument, that developers have had time to update, that the architecture move was necessary, that Rosetta bought people years, is the one Apple's defenders reach for immediately. And to be precise about this: the argument is not entirely wrong. Architectures do change. The issue is not the transition itself.

The issue is what you are paying for when you buy the hardware.

The last MacBook Pro with user-upgradeable RAM was sold in 2012 (the same year we formally launched Art of FACELESS). Since the Retina models introduced that year, RAM has been soldered to the logic board. Not because of a technical necessity that could not be solved any other way, but because Apple decided soldering was preferable and because soldering means the machine you buy today is the machine you keep until you buy another machine. The storage followed. Proprietary interfaces were introduced, for which no third-party alternatives were developed. The SSD that holds your work, your archives, your project files: non-replaceable.

A MacBook Pro with enough RAM and processing power to render complex 3D animation, the kind of machine a game developer, an animator, a motion graphics artist (that's some of us here at AOF) actually needs, starts at over £2,000. You cannot upgrade it. You cannot repair it cost-effectively outside of Apple's own service network. When the OS that supports your software is deprecated, your options are: buy a new machine, or stay on an ageing OS and accept the security risk.

This is not an accident.

It is a business model.

The Escape Route They Closed

For years, Boot Camp offered a partial answer. If macOS deprecated your software and, you could dual-boot into Windows and keep working. It was not elegant, but it was a pressure valve. A creative professional could justify the premium hardware cost partly on the basis that the machine ran two operating systems, giving it flexibility that a Windows laptop did not. We did exactly that. It worked.

Boot Camp does not exist on Apple silicon. It was dropped with the M1 chip in 2020 and has not returned. Apple silicon runs on ARM architecture. Windows for ARM exists, but Microsoft licenses it only to OEMs and provides no support for installation on Apple hardware. Apple provides no drivers. There is no dual-boot option, native or otherwise. The pressure valve is gone.

The alternatives offered are virtualisation tools — Parallels, VMware Fusion — which carry their own subscription costs and deliver materially worse GPU performance than native hardware access. For anything compute-intensive, they are not a substitute. For game development, 3D rendering, or GPU-dependent creative work, they are not remotely adequate.

So you cannot upgrade the hardware. You cannot run a fallback OS natively. And when the compatibility layer that kept your software alive is removed, you are told to update your apps or find alternatives. Some alternatives do not exist. Some updates cost more than the original software. Some software was owned by companies that no longer trade.

The Linux Question Nobody Wants to Answer

There is an obvious question hanging over all of this, and it has been hanging since the late 1990s. If macOS is too expensive, too locked-down, and too architecturally unstable for long-term creative workflows and if Windows is a bloated, telemetry-laden subscription service dressed up as an operating system, why is there no viable Linux alternative for creative professionals?

Linux distributions routinely support hardware that is fifteen to twenty years old. The same kernel that runs on a server rack will run on a machine from 2005. Nobody is deprecating your architecture because their silicon division has a new product to sell. Nobody is soldering your RAM to ensure you buy the next generation.

And yet. In twenty-five years of Linux development, with the full weight of open-source community effort and the financial backing of companies whose products depend on it, there is still no production-ready creative pipeline for the kind of work that demands it. No robust, stable, widely-supported equivalent of the tools that creative professionals actually use. Fragments exist. Workarounds exist. The pipeline does not.

This is worth sitting with. It is worth asking who benefits from that vacuum remaining unfilled. It is worth asking whether the absence of a credible Linux creative workstation is a failure of the open-source community, or whether it is something more structural, a market shaped by the interests of the companies who profit from the trap remaining sprung.

What This Is Actually About

The Fuck-AI crowd is loud right now. Fair enough, there are real concerns in that conversation, and AOF has documented and challenged specific failures of AI classification systems at institutional level. We know what unvalidated detection looks like. We know what it costs.

But here is what we have noticed. The same people posting furiously about AI from their iPhones, their Apple silicon MacBooks, their Windows 11 machines, the same people performing outrage about corporate technology on corporate technology platforms, are not posting about this. Not about the macOS 28 cliff. Not about Catalina's software graveyard. Not about £2,000 machines you cannot repair. Not about Boot Camp's removal. Not about the Linux creative desert that has persisted for a generation.

This is not new. This is twenty-five years of the same pattern, running uninterrupted, because the outrage always lands somewhere convenient and never quite reaches the business model.

Apple is a trillion-dollar company. The hardware costs what it costs. The OS deprecation runs on a cycle. The software graveyard grows with every transition. And the ecosystem is designed, precisely, deliberately designed, to make leaving more expensive than staying.

EOL is not a technical term. It is a policy.

This means that from a technical and creative's POV Apple, a trillion-dollar company, is an epicFAIL# candidate.

So, is this post a veiled dig at corporate power and capitalism?

No.

There's nothing veiled about it.

References

Apple Rosetta / macOS Transitions

• Using Intel-based Apps on a Mac with Apple Silicon — Apple Support — support.apple.com
• Apple Rosetta 2 EOL — macOS Sonoma Release Notes — OpenCore Legacy Patcher — dortania.github.io
• Apple Rosetta and macOS Ventura — OpenCore Legacy Patcher — dortania.github.io

macOS Catalina / 32-bit App Cull

• 32-bit App Compatibility with macOS — Apple Support — support.apple.com
• Apple Is Killing Off 235 Mac Apps When It Launches macOS Catalina — TechRadar — techradar.com
• 32-bit Apps No Longer Working on macOS Catalina — MacRumors — macrumors.com
• List of 32-bit macOS Apps — RoaringApps — roaringapps.com

Boot Camp / Apple Silicon

• Here's Why the Future of Boot Camp on macOS Looks So Bad — GroovyPost — groovypost.com
• Can You Run Windows Software on an M1 or M2 Mac? — How-To Geek — howtogeek.com
• Apple Should Bring Back Boot Camp — How-To Geek — howtogeek.com

OpenCore Legacy Patcher

• OpenCore Legacy Patcher — Official Documentation & FAQ — dortania.github.io
• Supported Models — OpenCore Legacy Patcher — dortania.github.io
• How to Install macOS on an Unsupported Mac — Macworld — macworld.com

epicFAIL#: It's so bad you can't tag it.

Art of FACELESS is an independent multimedia collective based in Cardiff, Wales, operating since 2010. We carry no third-party ads and no sponsors. Running independent infrastructure costs real money. If the work means something to you, please support independant publishing and the open web. You don't need any billionaire's platform to sign up or sign in.