Internet enthusiasts of the previous century sometimes expressed the power of code by declaring the sovereignty of cyberspace, or that "code is law."

As odd as these claims sound today, they hit a deep truth: end-users, and even governments, lack power to re-litigate choices embodied in software. Vendors, therefore, have power over them. Backed by deeply embedded control chokepoints, and without a proportional response from other interests, this control is akin to state power.

Both fear and fervour about these properties developed against a backdrop of libertarian¹ attitudes toward regulation and competition. Attenuating vendor power through interoperability was, among other values, a shared foundation of collaboration for internet pioneers.

The most fervent commitment of this strain was faith in markets to sort out information distribution problems through pricing signals,² and that view became embedded deeply into the internet's governance mechanisms.³ If competition does not function, neither do standards.

The internet's most consequential designs took competitive markets as granted. Many participants believed hardware and software markets would (or should) continue to decouple; that it would be ever easier for end-users to bring software of their choosing to devices they had purchased. Their faith in the trajectory was well-founded. It is ludicrous from the perspective of 2025 to suggest that swapping implementations of nearly any internet standard should come at the cost of replacing hardware.

Internet standards bodies assumed the properties of open operating systems and low-cost replacement of software to such an extent that their founding documents scarcely bother to mention them.⁴ Only later did statements of shared values see fit to make the subtext clear.

And it has worked. Internet standards have facilitated interoperability that has blunted lock-in, outsized pricing power, and monopolistic abuses. This role is the entire point of standards at a societal level, and the primary reason that competition law carves out space for competitors to collaborate in developing them.⁵

But standardisation is not purely an economic project. Standards attenuate the power of firms that might seek to arrogate code's privileges. Functional interoperability enables competition, and in so doing, reallocates power to users. Interoperability, and the standards that ensure it, are therefore at least partially a political project; one that aligns with the values of open societies:

We must ask whether ... we should not prepare for the worst leaders, and hope for the best. But this leads to a new approach to the problem of politics, for it forces us to replace the question: "Who should rule?" by the new question: "How can we so organize political institutions that bad or incompetent rulers can be prevented from doing too much damage?"

— Karl Popper,
"The Open Society and Its Enemies"

Without a counterweight, network effects allow successful tech firms to concentrate wealth and political influence. This power allows them to degrade potential competitive challenges, enabling rent extraction for services that would otherwise be commodities. This mechanism operates through (often legalised) corruption of both legal and electoral systems, and when left to fester, is caustic to democracy itself.⁶

As we shall see, Apple has deftly used a false cloak of security and privacy to move the internet, and web in particular, toward enclosure and irrelevance. Below, I develop the case for why Apple should be considered a corrupted, and indeed incompetent, autocrat in the digital lives of millions, abusing a unique form of monopoly to extract rents, including on the last remnants of open ecosystems it tolerates.

Worse, Apple's centralisation through the App Store entrenches the positions of peer big tech firms, harming the prospects of competitors in turn. I submit that Apple have been, over the course of many years, poisonous to internet standards and the moral commitments embodied in that grand project.⁷

Despite near continuous horse-race coverage in the tech press, the consequences of this regression in civic/technical affairs is not well socialised.

The Power of Interoperability

One reason to champion standardisation is that it signifies the expansion of interoperable technology, pushing firms to innovate, rather than extracting rents for access to commodity features.

Interoperability-in-being allows users to choose alternative implementations, forcing competitors to differentiate on quality and not yet standard features. Standards accelerate the emergence of true interoperability by lowering the costs of implementation and reducing tails risks to implementers, e.g. from patent trolls. Over time, a vibrant and growing set of standards can attenuate the power of vendors to extract rents and prevent progress in important domains.

Interoperability is not the only mechanism that can dampen the power of dominant firms, but it is the most powerful. Free and Open Source software (FOSS) can provide a counterweight to rentiers, but OSS is not a full solution.⁸

Voluntary Adoption Is Foundational to Internet Standards

Interoperability, and the economic surpluses that flow from it, are underpinned by the bedrock principle of voluntary adoption. This is enshrined in the "open stand" principles agreed to by no less than ISOC, IETF, IAB, IEEE, and the W3C.:

...

5. Voluntary Adoption

Standards are voluntarily adopted and success is determined by the market.

— The IAB, IEEE, IETF, ISOC, and W3C,
"The OpenStand Principles"

This final principle is the shortest, and many readers will understand it as a dodge; a way for Standards Development Organisations (SDOs) to avoid being seen to "picking winners". But it is not only that.

Voluntary adoption is necessary for internet standards to function, and this principle implicitly binds participants to a presumption of fair play, both for themselves, and for others in the ecosystems that standards bodies facilitate.

Several implications bear mentioning.

First, the principle of voluntary adoption is a predicate for effective standards' development. Without some mechanism for determining which designs are good vs. bad, we are unable to make consistent progress, and that test never comes from within an SDO; it is always extrinsic and customer-defined. Writing a standard is not a test of quality, and conversely, without a functional market in which to test designs, SDOs are irrelevant.

Second, this principle outlines a live-and-let-live doctrine, both within standards bodies and in the market. Participants may want their design to "win", but are enjoined from using procedural shenanigans to prevent competing designs from also being standardised.

Lastly, voluntary adoption marks customers (developers) and suppliers (browser vendors, etc.) as peers worthy of mutual respect and creates a clear norm within the walls of SDOs.

For all of these reasons, voluntary adoption must be inviolable, and actions taken to undermine it must be met with resistance and eventual sanction.

Apple's Unique Monopoly

Regulators have had no difficulty in building market tests that demonstrate the power Apple holds in the lives of users.

Some of these tests have produced comical contortions as Apple has attempted to weasel out of its responsibilities. Consider the Trinitarian claim that Safari is simultaneously one product, and also three. Or that iPadOS, despite sharing nearly all code with iOS, being marketed under that brand for many years, supporting considerably identical features, running all the same third-party software, and being released on substantially the same cadence, while running exclusively on hardware of the same architecture as iOS devices is, somehow, an entirely different product.⁹

But these tests, for as clear and effective as they have been, do not capture the most important aspects of Apple's influence on the market for smartphone software.

Apple's ability to negatively impact the potential of standards-based platforms to disrupt the App Store relies on properties of the market that every software developer understands: a monopoly on wealthy users, and through it, influence.

Competition law does not explicitly recognise or endorse this distortion of the market structure, but the connection between the influence of wealthy users and the downstream choices available to other end-users is indisputable in mobile ecosystems.

Despite selling fewer than a quarter of smartphone devices every year for the past decade, Apple has maintained iron control over the way smartphone software is developed and delivered, owing to the social and market effects of wealthy users gravitating to iOS devices. Apple uses both legitimately superior product attributes (e.g., leading chip design) along with anti-user and anticompetitive tactics — e.g., green vs. blue chat bubbles, suppression of browsers — to maintain this position.

Apple imposes an interlocking set of restrictions to help ensure that standards-based platforms cannot disrupt its position as the primary arbiter and delivery channel of software for wealthy users. Those users, in turn, enjoy disproportionate influence over the behaviour of software developers, owing both to their greater spending power, and their collective ability to determine the relevance of competing platforms thanks to their positions within the software industry. Developers of all stripes understand that if they cannot demo their wares to the bosses and VCs (all of whom are wealthy) on their own devices, their software might as well not exist.

The long-stable propensity of users who make more than $100K USD/year to carry iPhones combines with Apple's suppression of browsers and PWA capabilities to ensure that developers have no effective choice but to build native applications, for which Apple extracts rents. These effects were visible in population-level statistics a decade ago and have been stable ever since:

Knowing whether someone owns an iPad in 2016 allows us to guess correctly whether the person is in the top or bottom income quartile 69 percent of the time. Across all years in our data, no individual brand is as predictive of being high-income as owning an Apple iPhone in 2016.

— Marianne Bertrand and Emir Kamenica,
"Coming Apart? Cultural Distances in the United States Over Time"

Over time, the effect becomes self re-enforcing: developers are forced into the App Store due to a lack of capabilities in browsers, granting proprietary ecosystems a library of software with no interoperable parallel. This further induces wealthy and influential users to seek software exclusively through App Stores.

The monopoly on influence explains in part why Apple is wedded to legalistic, dissembling tactics in order to prevent the spread of standards-based platforms. Should the work of internet and web standards bodies ever become relevant, Cupertino understands the market for software will transform in ways it cannot control or tax.

How Apple Uses Its Monopoly to Centralise and Enclose

Developers are forced into the App Store, first and foremost, by a lack of functionality in standards-based alternatives. By contrast, Apple has used over-broad grants of system capabilities to woo developers with an interest in monetising user data through too-broad grants of detailed information about users and unconscionable allowances for suppressing pro-privacy interventions by the most rapacious native apps.¹⁰ When Apple professes that "privacy is a human right", we must understand this as an attempt to turn the consequences of Apple's own largesse towards data abusers into a marketing asset and an attempt to convince users not to flee the proprietary ecosystem it favours.

These over-broad grants depend, fundamentally, on Apple's own decisions.

It was Apple's choice to introduce less safe, less privacy-preserving native apps into iOS. It was also Apple's choice to deny competing browsers engines the freedom of voluntary feature adoption, ensuring that important underlying capabilities could only ever be accessed through Apple's proprietary APIs, and only ever by those who are willing to agree to Cupertino's terms.

The result has been API enclosure; appropriation of commodity capabilities that themselves are standards-based — e.g., USB, Bluetooth, NFC, file storage, etc. — by a proprietary ecosystem and denial of even the safest and most privacy-preserving versions of those features by open, interoperable, and standards-based application platforms.

This, in turn, has created a winner-take-all dynamic inside the app store, harming privacy, security, and competition in the process.

Apple's Transgressions Against Voluntary Adoption

This strategy relies on a set of interlocking policies that harm competitors and prospective challengers. The sabotage of voluntary adoption lies at its heart. A Bill of Particulars for crimes against the web and internet community springs from a relatively small set of undisputed facts.

Apple has:

• Restricted competitors from delivering their own implementations of web and internet standards, depriving them access to Apple's monopolised controls of influential users.
• Forced all iOS browsers to use Apple's own, defective and impoverished, implementations, depriving users of choice and destroying the market's ability to select for better ideas and implementations.
• Compelled browser engine monoculture and, in so doing, consistently undermined user security and privacy.
• Attempted to use contractual terms to dissuade competitors from extending browsers to support standards that Apple disfavours.
• Serially misled regulators and the public when presented with evidence of the harms that spring directly from the above.
• Objected spuriously in bad faith within standards bodies to prevent standardisation of features which Apple offered no counter-proposal.¹¹
• Since at least 2015, delivered explicitly pejorative UI and marketing to discourage use of open and interoperable technologies that might benefit users at its own expense.

Through these and other overt acts, Apple has worked to disempower users, depriving them of choice in the market, and thereby devaluing the effectiveness of interoperable, standards-based platforms.¹²

These acts are not simply the competitive acts of a fierce market participant. Apple has done violence to the founding ethos of internet and web standards development. Instead of honourably withdrawing from those groups, Apple has maintained a charade of engagement, and gaslights other participants while actively sabotaging the principle of voluntary adoption that internet standards are predicated on.

Unilateral Off-Ramps

It must be emphasised that Apple has never been forced to suppress its competitors, nor to create an anticompetitive landscape. At every moment, Cupertino's senior management have retained intellectually consistent choices that allow it to pursue growth of Apple's superior (we are told) native app ecosystem without creating conditions that threaten browser choice or the good functioning of internet standards.

Let's consider two: all safe browsers, and no browsers.

Apple could, of course, simply enable the same sort of level playing field for high-quality browsers that every competing vendor has for nearly the web's entire history, and which Apple has itself facilitated on macOS.¹³ Any plausible restriction owing to available system resources has long been overcome by progress in mobile hardware, particularly within Apple's ecosystem.¹⁴ Early mobile era justifications based on API richness and resource availability were falsified by even the lowliest Androids to more than a decade ago.

The only conscionable restrictions on competing browser engines spring from demonstrable failures to safeguard security. As other vendors have generally had better track records vs. Apple regarding sandboxing strength, security response times, patch gap width, and consistent coverage for users on older OS versions, this should be no practical obstacle.

Lest Apple's defenders worry about the impacts on the company, under true browser choice, Apple retains considerable market advantages, including (but not limited to) pre-installation, lower structural costs¹⁵, and continued voluntary feature adoption, allowing it to earn users through differentiation.¹⁶ Such bulwarks have allowed Safari to retain considerable share on macOS in spite of stiff competition and Safari's poor track record on security and standards conformance.

Alternatively, Apple could withdraw Safari while forbidding all browsers on iOS. This is a fully consistent position, and one that has been available to Apple from the moment of the iPhone's release. The iPod did not include a browser, and many subsequent Apple OSes lack functional browsers. iOS and VisionOS are uniquely deficient in this regard.¹⁷

Either way, the choice to undermine choice and standards has rested entirely with Apple. At every moment it has had intellectually honest solutions available to resolve these contradictions of its own design. Apple cannot claim the situation is anyone else's fault, or that it has had no choice.

How Apples Crimes Differ From Prior Episodes

It should be obvious from the proceeding sections that I do not consider Apple's mere failure to implement standards I personally approve of to rise to the level of the transgression worthy of sanction.

Under voluntary implementation, every vendor is free to ship whatever they please, including Apple. Whilst it may be sad, or even damaging, when features go missing from important products, that is not a calamity. In functioning markets, this is simply an input to be priced. Among browsers, this has traditionally been experienced as vendors gaining or losing share to the extent they support otherwise-interoperable content.

And so it is not enough to cite a lack of features, bungled implementations, peevish behaviour in working groups — or even rank dishonesty — as reason for censure. These are, to a greater or lesser degree, players playing the game within the rules. Some tactics may be distasteful, but reside squarely within the "awful but lawful" category. Venues dedicated to free exchanges of views should allow them, with sanctions for poor behaviour meted out in the social realm, if at all.

Apple's outrages against this community are more fundamental, and more dangerous.¹⁸

Some will see here a parallel to the Paradox of Tolerance, and I do not believe this is mistaken. Standards bodies can, and should, admit many positions by their participants, but bestowing membership in good standing to those actively uprooting the basis for standards is to ensure that standards, and the ecosystems that depend on them, wither and die.

By subverting the voluntary nature of open standards, Apple has defanged them as tools that users might use against the totalising power of native apps in their digital lives. This high-modernist approach is antithetical to the foundational commitments of internet standards bodies and, over time, erode them.

Indeed, no other vendor has achieved what Apple has in the realm of anticompetitive suppression of the web. We must not imagine that Apple would stop at the Application layer given a chance. The same mechanism threatens voluntary feature adoption in every other layer of the stack, too.

Necessary, Proportional Responses

The web and internet communities must understand not only the threat, but the ongoing harms to the cause of the web and of internet standards. It seems to me that this point has hardly been engaged, let alone won, within the walls of SDOs. But if it were, then what? What, practically, can be done?

The founding documents of internet SDOs do not include mechanisms for censure of these violative acts. The W3C's bylaws, for example, only speak to membership in good standing in relationship to payment of dues. Regardless, it is possible — and I believe urgent — to do more.

First, proposals can be raised to amend those documents to include mechanisms for censure by votes of the membership for actions such as those alleged here. These are likely to fail, and will surely be rejected on a first attempt, but the act of constituents considering these questions has power in and of itself. Raising proposals for discussion at plenary meetings and in visible fora can, at a minimum, elicit a response to the charges levied here. That, on its own, is valuable to the community.

Next, leadership boards with moral authority can write persuasively on the question. The W3C's Advisory Board and Technical Architecture Group and the IETF's Internet Architecture Board have the ear of the membership, even on non-technical topics, should they choose to weigh in on the side of the continued relevance of their convening organisations.

Most importantly, individual delegates to working groups of all sorts can recognise that Apple's compelled use of WebKit by competitors on iOS is illegitimate and corrosive. Having done so, they can resolve not to accept "Apple does not comment on future products" as a viable response to questions about implementation timelines.

As long as Cupertino demands and unjustly defends a monopoly, it is incumbent to demand responsibility for the consequences of Apple's failures.

Apple alone can choose to support features within WebKit and allow them in other iOS browsers, even under compellence, and even if it does not to enable them in Safari. It is simply illegitimate for Apple to claim that it cannot or should not allow other vendors to reach feature parity with competing engines.

The sham of WebKit as an Open Source project is incompatible with disavowal of features that other vendors would flag on for their iOS users if allowed. Compelled implementation and the destruction of voluntary adoption should not be a shield against critique. Instead, it must heighten expectations. These are simply the consequences of accepting responsibilities that Apple agues it should be singularly entrusted with, despite Safari and WebKit's trailing record on standards' conformance, security, and privacy.

Apple alone must be on the hook to implement any and every web platform feature shipped by any and every other engine. It does not need to enable them in Safari, but must make them available for use by others as they see fit.

So long as competing vendors are forced into the App Store and must use Apple's engine, Apple should bear the costs of completeness and feature quality. So long as Cupertino compels use of WebKit by competitors, the demand must be echoed back: parity with browser features on other Operating Systems is the minimum bar.

WebKit purports to be Open Source, but in practice Apple has used it to undermine the "bring your own code" foundation of OSS. It is illogical for Apple to cite a disinterest in a feature in Safari as a reason for Apple not to be expected to implement those features in iOS's WebKit binary, making them available for other embedders to flag on.

Fundamentally, the web and internet community must stop accepting the premise that Apple should benefit from the protections and privileges that voluntary adoption affords whilst it denies those benefits to others.¹⁹

Lastly, and perhaps most controversially, delegates and organisations can use their positions to vote against Apple's personnel in elections to leadership positions within internet and web SDOs. It is, of course, inconsistent for Apple to hold positions of influence in organisations they are actively suppressing, and fellow participants are under no obligation to pretend otherwise or hand Apple formal or even persuasive power within these groups.

Why Now?

In raising these questions, colleagues have invariably asked "why now? What changed?"

Beyond the threshold point that the damage done is cumulative, and therefore it isn't necessary to identify specific instances to discuss the rot Apple's influence has caused, it's fair to ask why anyone should be agitated tomorrow when they might not have been yesterday.

Most of the factors involved have indeed changed very gradually, and humans are famously poor judges of slowly emergent threats. Apple's monopoly on influence, Cupertino's post-2009 WebKit priorities, the suppression of browser competitors, and the never-ending parade of showstopping bugs are all gradually emergent factors. Despite all of this long-running, unrefuted evidence, many continue to think of Apple an ally of the web for helpful acts now more than 15 years old.

But recent events must shake us awake. Apple's petulant attempt to duck regulations, destroy the web as a competitor for good, and frame regulators for the dirty deed was shocking. In recurring misrepresentations to regulators before and since, it has dissembled about its role in suppressing the web, and through its demand for secrecy in quasi-standards processes, has worked tirelessly to cover its tracks.

Taken individually, and in ignorance of iOS's coerced WebKit use by competing browsers, forced monoculture, habitual security failures, and strategic starvation of the Safari team, these shameful acts would simply indicate another monopolist behaving badly. It is only when considered alongside the wider set of facts that the anti-standards strategy and impact become clear.

Do Standards Matter?

Like most who have dedicated the greater proportion of their working lives to the cause of an open and interoperable web, the conclusions offered shock me as well.

In the end, however, the question "do standards matter any more?" is intrusive, despite my aversion to reconsidering a question whose answer I thought obvious. But in light of the past decade's sidelining of the web, we must grapple with the consequences. To allow Apple to continue to abuse the foundation of standards without acknowledgement would be a failure of honesty towards my own intellectual commitments.

My personal affinity for the many talented and thoughtful people that Apple has sent to standards bodies over the years — including those I think of as friends — has, on reflection, been an emotional blind spot. But here we are. The realisation that they have been an unwitting fifth column against the web is nauseating for me, and I expect many will loudly reject the conclusion. I do not blame them.

For all the harm Apple has done to the web and to competition, I had hoped that it would relent before any of this became necessary. Like most web developers, I harboured hope that, true to Steve Job's promise in '07, Apple would let the web be a "sweet solution" for delivering safe, powerful applications. Piggybacked on that hope was a belief that a relevant mobile web would bolster the relevance of standards. But Cupertino has gone a different way, choosing profit over collaboration and the needs of users.

The web matters too much for the standards-based future it represents to fade without so much as a nod.

FOOTNOTES

1. Or, if you like, "liberaltarian". ⇐

2. A faith I do not share. Markets fail frequently, never mind that most goods are hardly substitutable, thinly traded, and lack reliable public prices. All of this means that the information capacity of markets a priori the rationalising effect of standards and market fairness regulation is heavily suspect.

   But even for those that take a market fundamentalist perspective, restrictions on trade such as Apple has imposed are offensive to the basic logic of the market's role in bettering of society. Only those who would see markets subjugating all, forever, because of one-time power imbalances can be sanguine about what Apple has done.

   We do not have to grant that pro-totalitarian arguments in technology are in good faith given what we know about where unchecked power reliably leads. ⇐

3. It is not original to note that there is an inherent contradiction in the idea of liber(al)tarian participants in standards bodies collaborating through non-market mechanisms. We do not have to ignore it, however, or even treat those holding both an affinity to open standards and libertarian ideals as hypocrites, in order to accept that the development of standards is often a creative act of critique for which there is no other functional venue (or, if you like, "market").

   Proposals for open internet standards often begin as personal drafts of individual authors, working in a community of like-minded developers experiencing similar challenges and working to design solutions. These proposals are situated in a context that is often opaque to those outside narrow circles, and the communities that form around them trade in reputation as much as any other currency.

   In addition to personal status, there is a distinguishable ideology at work:

   The open systems ideology that was developed in computing between the 1970s and the 1990s embodied several assumptions articulated in previous open system visions in diplomacy, economics, philosophy, and engineering. These assumptions included:

   • an economic commitment to global markets;
   • a moral support of international and multicultural ties;
   • a political opposition to centralized power — either in governments or in monopolies — that threatened individual autonomy;
   • a belief that technical professionals could achieve these economic, moral, and political aspirations through cooperation and standardization.

   — Andrew L. Russell,
   "Open Standards and the Digital Age"

   These principles stand as critique to older, less open ways of working. Technology that is more interoperable makes a larger space for critique through code and the market, and that is essential to the openness of any society that depends on software as much as modern, western nations do today. ⇐

4. The history of Bell Labs, the antitrust battles of IBM, the birth of Unix, and the expectations of "common carrier" treatment for data transmission help to situate the thinking of participants in setting up the foundational standards bodies that we now take for granted. Major battles were fought to keep computing out of the hands of a few corporations with outsized power, and the consequences continue to reverberate.

   It is my personal view that it has been the good work of antitrust and anti-monopoly reformers — both within and outside government — that made it possible for other parties to believe in the abstraction of functional markets. Functioning markets are not, in fact, self-organising, and are enabled explicitly by law that helps to minimise noise in the channel (e.g., from fraud).

   That self-identified libertarians (along with the rest of society) have benefited dramatically from the gains enabled by these anti-monopoly regimes is not in doubt; the only mystery is how fervently some cling to the belief that regulation must be problematic as a way to skip past its content and ignore engaging on substance. ⇐

5. Without an "all clear" signal for explicit standardisation, antitrust law in most advanced nations would explicitly forbid the sorts of coordination among industry peers that standards bodies facilitate. Authorities permit, and even encourage, standardisation in contrast to other forms of collaboration because the positive externality benefits of reduced friction to trade from interoperability is so compelling. For more on the history and structure of this now-international regime, see Russell (2014) ⇐

6. One needs to look no further than tech CEOs lining up to hail the rise of explicitly fascist leaders and ply them with gifts, to understand the ways in which this corruption corrodes our hopes for open, tolerant societies. ⇐

7. This is not an accusation to be levelled lightly, and not without overwhelming evidence. But that evidence has accrued, and so I feel compelled to speak.

   Apple are known both to be capricious and retributive, and its antipathy towards Khronos provides an excellent example of the decade-length grudge-holding for which Cuptertino is infamous. I therefore write this post advisedly and with trepedation. ⇐

8. More than twenty years into the experience of Open Source as a force in software, we can clearly discern that the licence of source code is not, in-and-of itself, a solution to the imbalances inherent in software's relationship to users, or even to other parties with the capacity to develop software.

   I believe that history demonstrates clearly that (F)OSS licensing and effective governance are only related by the intellectual and commercial commitments of those building software, and are therefore easily co-opted by wealthy and powerful firms.

   We have witnessed the limits of forking as a bulwark against bad behaviour, through both the persistent upward reach of firmware into "open" systems, and through capture in higher layers due to the carrying costs of complex systems. Licensing has also been little help in courts against large and determined enemies of OSS and the freedoms it attempts preserve. The limits of licensing, and the lack of IP defence pooling inherent in common licences, create risks that successful firms and projects hit regularly.

   At the limit, (F)OSS licensing is not a significantly disruptive force against wealthy and powerful technology firms, but rather a tool that is useful to peer-level adversaries. For OSS licences and defences to have purchase in practice, a significantly endowed group of technologist's interests must be at stake. The interests of informed and financially capable technologists and the firms that use their software bear only passing resemblance to the needs and interests of society more broadly.

   From this perspective, we can understand (F)OSS as complementary to standards development, but unable to fully replace standards-based interoperability as an attenuating force on the power of software in the lives of users. Standards retain a unique ability to build space for challengers within and between proprietary products, which (F)OSS do not. The power of SDOs to pool the patent interests of proprietary players has parallels in (F)OSS, but with more clarity and lower risk, further accelerating practical interoperability.

   Therefore, we must understand that it is a rhetorical and intellectual trap to consider (F)OSS a replacement for standards. Each hasten interoperability and attenuate power in different and complementary — but not substitutable — ways, and both mechanisms are healthier when the other succeeds. ⇐

9. The EC, having borne witness to Apple's appalling behaviour within its borders, and more generally to boot, was having none of it.

   The full finding (PDF) is a masterclass in even-handed consideration, and as such it is not surprising that Apple's legal and marketing fictions failed. ⇐

10. Helpfully for Apple, the conversation around privacy and technology has barely progressed past Apple's anti-Google and anti-Facebook kayfabes.

    Privacy advocates are regularly taken in by Apple's marketing, and the tech press remains in a largely stenographic mode. None of this is to say that Google or Facebook are good actors (they are not), only that Apple is also not on the level.

    We can begin an analysis by noting that Apple does not encourage users to move their computing to the web where browsers can attenuate the worst invasions of privacy. Browsers do not provide nearly as much information to trackers as even the most locked-down native app in Apple's ecosystem, and Apple know this.

    Next, Apple has not funded lobbying and regulatory outreach efforts to establish privacy laws worth a damn. Instead, it channels huge amounts into defeating right-to-repair legislation and defeating browser engine choice around the world, including going as far as funding astroturf groups to provide Cupertino's views in stereo.

    Moreover, Apple takes no responsibility for its historical role in the growth of tracking via the native API surfaces Apple created relative to the web. Nor does it demand audits of data use by App Store participants, or to even set policy about acceptable uses of data collected via App Store-vended applications.

    Apple does not even forbid pervasive "ad blocker blocking" by the worst actors in its ecosystem, even when interventions are trivial from a policy and technical perspective.

    That all of this aligns with Apple's preference for control over, and taxation of, developers cannot escape comment. It further forces us to entertain the idea that Apple's position as a defender of privacy is a cynical show, or that it is incompetent in assessing privacy risks, or that it is out of touch with the behaviour of snooping developers. Any of these would be enough to judge it an incompetent regulator, asleep at the switch.

    Depressingly, these are not even exclusive choices. ⇐

11. For a vendor as wealthy as Apple, and one that insists it must maintain a monopoly on implementations of standards-based platforms on the most influential OS — particularly one that skims as much cash from the open web as Apple does — to object to proposals other vendors ship safely for years, without offering their own designs to address similar needs, is prima facie evidence of bad faith.

    This case is bolstered by implementation of the same designs to which spurious and unbacked objections were previously raised as regulatory pressure has grown. ⇐

12. Platforms are in competition.

    It cannot escape notice that Apple's undermining of voluntary adoption most damages interoperable platforms that might challenge the App Store-based native app ecosystem which supports Apple's vertical integration agenda. ⇐

13. Here we must mention other instances of operating systems imperilling browser choice. Microsoft, for instance, was credibly accused of making text less readable on Netscape Navigator.

    Google, for its part, introduced ChromeOS without any provision for competing browsers. Subsequently, it allowed Android versions of competing browser products using their own engines to register as the default system browser. This is an unsatisfying solution, as Play-based apps are a poor experience on low-end devices and competing browsers are prevented from integrating as deeply as Chrome does. This may not rise to the level of Apple or Microsoft's acts as ChromeOS remains a niche product, ranks lowest in influence, and features a capable browser — regardless, the precedent, combined with prior episodes of Microsoft and Apple turning away from their successful and capable browsers, remains troubling. ⇐

14. No company in the world is more aware of the awesome power and capability of Apple's A-series chips than Apple itself.

    Apple executives know, for instance, that even the least expensive iPhone produced in the past five years is hundreds of times faster than the first iPhones which gave rise to resource-based restrictions on engine choice. They are either aware, or can trivially deduce, that there is no legitimate system health or resource-related reason to disallow competing browser engines. ⇐

15. As I have discussed before, Apple economises on the development of WebKit and Safari not only by failing to fund development of web features in a timely way, drafting on the path-breaking work of others, but also through an (over)reliance on OS components in lieu of more easily defended abstractions. This goes much deeper than only needing to support a single family of operating systems; unlike competitors, Apple directly leans on OS systems where competitors rebuild large swaths of the runtime to enable the web where underlying OS APIs are more limited.

    The result, for Apple, is reduced headcount and a lower bill-of-materials when producing devices (read: higher profits), owing to higher code page sharing across applications.

    The consequence for users is an implementation monoculture and slower patch delivery, harming security. Apple's users remain vulnerable to attacks for longer and without recourse to competing alternatives with stronger track records. The downside of all high modernism is brittleness, borne of uniformity, and iOS is no exception. The lack of the ecological diversity that Apple demands undermines security and resilience. This is not a price society should pay for the convenience of a blue chat bubble. ⇐

16. In testimony before various competition authorities around the world, and in peevish screeds in response to those arguments failing, Apple habitually attempts a rhetorical redirection that unmasks its anti-Popperian thirst for control.

    It appears to believe, on the strength of brand, that it can win on the ground of "who should rule?," rather than "how can institutions keep bad rulers from doing too much damage?" In so doing, Apple offers an authoritarian model of technology. Cupertino proposes that all sources of control attenuation are themselves totalising, and any compromise forced on the ruler therefore equivalent to a coup.

    This is nonsense, both practically and historically, and technologists that support democratic norms should reject this framing.

    History shows clearly that open systems in open societies enable civil society to take many effective positions that attenuate untrammelled power, both of the state and of other actors. We can further see that neither open nor closed systems are full bulwarks against government coercion. The prospects for improved societies must not be invested in autocratic technology firms for the simple reason that they are powerless to deliver it.

    Apple itself has been forced to compromise in areas such as right-to-repair without any of the apocalyptic prophecies of Cupertino's lobbyists coming to pass.

    Here Apple, and its extremely vocal band of apologists, will bring up overreaches by others, including disastrous and ill-advised anti-encryption pushes by national governments. The goal of this argument is to sully the idea of democratic control of technology. "How", it asks leadingly, "can anyone trust these people to do what's right for users?"

    How indeed! For the central point is that we do not have to. Recourse is available (in functioning democracies) through elected representatives and the responsiveness to citizen's concerns. We can even accept that democracies will get many of these issues wrong for a time without seceding any ground to Apple's authoritarian offer. It is intellectually consistent to reject both noxious positions taken by elected officials and offers of high modernist control by firms opposed to the current positions of a government.

    Apple's arguments are microns deep on the merits. They attempt to erase both the successes of civil society (rather than Apple) in curbing abuses along with Apple's own shameful track record of capitulation to overtly authoritarian regimes.²⁰

    Apple's framing asserts that total control by a famously unaccountable firm is a Panglossian utopia. That any choice Cupertino makes now, has ever made, or has ever reversed, have unquestionably been the best of all possible choices. That by paying a thousand dollars for a phone, we are but lucky to bask in the glow of such resplendent wisdom.

    It's all too much to take, at least for anyone with a working memory. As I hope I have demonstrated here, this posture fails on ethical as well as practical grounds.

    Apple, with all of its power and money, could be an ally to civil society in curbing abuses without claiming total and unaccountable control for itself in the process. It is incumbent on any thinking technologist to reject such land grabs, even when they come wrapped in causes we otherwise support. To do otherwise is to co-sign the doctrine of "who should rule?," a framing that is caustic to both our short-term technical and long-term societal interests. ⇐

17. Paradoxically, it may be Apple's own malign behaviour regarding browser choice that might prevent a "no browsers" policy from being possible today.

    Even if we were to ignore Cupertino's incentive to maintain browsers on iOS owing to the shockingly large flows of money from Google for search placement (which Apple records as nearly pure profit), regulators may require it to pursue an "all browsers" alternative to fairly redress the harms from the previous 15 years of abuses. Should this come to pass, Cupertino would (again) have no party to blame but itself. ⇐

18. The damage Apple has done to the cause of internet and web standardisation is analogous to the (US) concept of impeachable "high crimes and misdemeanours", rather than better enumerated, more pedestrian infractions. This category covers acts that are destructive to the foundational principles of the enterprise, regardless of narrow legality.

    Internet SDOs are not set up well to police or react to this class of offence, which may help explain why Apple's violence against the community has gone unremarked for so many years. ⇐

19. Apple engineers should be questioned (politely, but insistently) about timelines for implementation of any feature that any other engine chooses to ship, and failure to provide that timeline should be viewed as ongoing evidence of malfeasance against the internet and web community for as long as Apple withholds the ability for others to bring their own engines and whatever features they choose.

    If this causes Apple to retreat from sponsoring work in these venues, that is regrettable, but must also be understood as a choice that is entirely within Apple's power to reverse. ⇐

20. The fuller airdrop story creates a cross-pressured narrative in which it is possible to claim that Apple was acting to protect, rather than suppress, A4 protesters. But such a pose must also contend with Apple's own history of failure, first to build a truly private version of a system marketed in those terms, thereby encouraging users in sensitive situations to expose themselves to great risk. Second, Apple's unwillingness to patch post-disclosure.

    Owing to AirDrop's use of closed protocols that Apple does not make available to developers of competing applications, it also falls exclusively on Apple's shoulders that better solutions were not available to replace Apple's own botched implementation.

    The story is more complex than "Apple sold protesters up the river," but no publicly available version of events is a defence the argument that Apple is a bellicose marketing outfit with a situational and opportunistic relationship to user privacy and security.

    When the effect is to expose or debilitate users that rely on the very properties Apple so truculently asserts its competitors fail to deliver, it becomes impossible to defend Cupertino's insistence that it alone should be trusted. ⇐

When I still worked for a US company, I traveled every few weeks to the US, so much so that I had gold status on several airlines. Ever since the Covid pandemic hit and I got laid off by my last employer as part of a 10k employee cut, I stuck to Europe. Other reasons […]

A few weeks ago I participated in this explainer video for the UCSF Cancer Survivorship program and it's out now:

I think the video does a nice job of explaining the primary mechanism for my ongoing interactions with UCSF over the last few years (I mean, other than the colonosopies). It was also nice to see glimpses of how this program has already helped so many others navigate that post-treatment life re-integration stuff. I'm very lucky to be working with Angela (she's in the video!) on my own post-treatment re-integration stuff!

Last related tidbit - if you're into running and in the Bay Area this weekend, the Run Tiburon 5K / 10K road race is this coming Sunday, Sept. 7th (you can still sign up!) and a portion of proceeds go to benefit the UCSF Cancer Survivorship Program. They've even renamed the race "Courage Over Cancer." Somehow I came in 7th place in this race two years ago, which only means one thing - we clearly need more runners to join!

A design language, if you can keep it.

I’ve only been married for a few years, but I have a fantastic marriage tip you won’t hear from any marriage counselor or book:

• Get a couple’s email domain

What’s a couple’s email domain?

My wife and I share a .com domain name for email. I’m not going to reveal our real domain name, but pretend it’s this:

• @shinytable.com

Emails to michael@shinytable.com go to both me and my wife, and the same for her name.