Two projects, and neither of them was anybody's job

When Adobe ended a runtime, nobody at any studio owed the archive anything. Two volunteer efforts decided that was not good enough.

The problem Adobe left behind

On 31 December 2020, Adobe ended Flash Player support and the browser vendors followed within days, removing the plugin from their update channels entirely. The decision was defensible — Flash carried a long history of security vulnerabilities, and the open web had grown capable enough that the plugin's unique advantages had narrowed to near zero. But defensible is not the same as costless. Somewhere between seventy thousand and a hundred thousand browser games had been built in ActionScript over twenty-five years, and the vast majority of them had no other home. They were not on storefronts. They had not been ported. Their developers had moved on, or in many cases simply could not be found. The games existed as SWF files scattered across servers that were themselves being decommissioned, and without a runtime to interpret them, SWF files are inert.

No commercial actor had a clear incentive to fix this. Adobe had already made its call. The browser vendors were not in the preservation business. The games industry does not have a systematic archiving mandate the way film or recorded music industries have cautiously begun to develop. What remained was a question of whether anyone would act anyway — and two projects did, for different reasons and by different means.

Ruffle: the emulator built from the inside out

Ruffle is an open-source Flash Player emulator written in Rust, capable of running directly in the browser via WebAssembly without any plugin installation. The project predates the end-of-life announcement — work began around 2019 — but the December 2020 deadline gave it urgency and a sharp public profile almost overnight. The core insight was architectural: rather than trying to preserve the original binary and somehow keep it running, Ruffle reimplements the Flash runtime from scratch, reading the same SWF format and executing ActionScript 2 and, progressively, ActionScript 3.

That distinction matters enormously. A preserved original binary depends on the operating system and browser environment remaining compatible, which they will not, indefinitely. An emulator written in modern, maintained code can be updated as the web platform evolves. Ruffle runs where WebAssembly runs, which is everywhere a contemporary browser runs, and it requires nothing from the user. The Internet Archive integrated Ruffle into its software emulation layer, which is how millions of archived SWF files became playable again through a standard browser tab with no installation required.

The limitation is honest and the developers do not disguise it. ActionScript 3 support is substantially more complex than AS2, and some games that rely on particular Flash APIs or precise rendering quirks still do not work correctly. Ruffle is a living project maintained by volunteers: progress is real but uneven, and it depends entirely on contributors choosing to spend their time on it. There is no payroll, no commercial backer with a roadmap commitment, and no deadline that anyone is contractually obligated to meet.

That is precisely what makes it remarkable and precisely what makes its long-term continuity genuinely uncertain.

Flashpoint: the different problem

Ruffle addresses playability in the browser. Flashpoint addresses something adjacent but distinct: the fact that many games were never purely local files in the first place. They phoned home. They loaded assets from servers that have since been shut down, validated sessions against authentication endpoints that no longer exist, or required a connected world-state that was switched off years before Flash itself ended. A browser emulator running a broken SWF is not preservation — it is a broken game. Flashpoint, a project initiated by BlueMaxima and expanded by a large volunteer community, attacks this from a different angle.

The Flashpoint archive is a locally-run collection — a curator's selection of browser games preserved not just as files but as complete, working software environments. It packages the necessary runtimes, spoofs the server requests the games expect to receive, and serves assets from a local web server so that games which were designed to be online-dependent behave as if they still are. The archive covers Flash but also Shockwave, Java applets, Unity Web Player content, Silverlight games, and other formats that the open web quietly made obsolete alongside Flash. By the time the project had grown to tens of thousands of titles it had become one of the most significant collections of playable browser-game history in existence.

The scale came with legal complexity that the project has never fully resolved. Archiving software for preservation purposes sits in a complicated position in most copyright regimes. In the United States, Section 108 of the US Copyright Act provides certain exceptions for libraries and archives, but Flashpoint is not a formal institution. The Library of Congress has documented the tension between copyright duration and digital preservation for decades without the underlying statute changing enough to give volunteer archivists real legal clarity.

Flashpoint operates in that gap, prioritizing preservation over waiting for legal certainty that may never arrive.

What the two projects reveal about the ecosystem

Ruffle and Flashpoint share a premise: that the cultural record of browser gaming has value worth protecting, and that if institutions will not act, individuals will. What they also share is structural fragility. Neither has guaranteed funding. Both depend on contributors remaining interested, and on the underlying platforms they target not changing so drastically that the projects must be rebuilt from scratch.

South Korea, which was among the first countries to build a mass broadband infrastructure and where PC cafés made browser and online gaming a genuinely mainstream culture years before the West caught up, saw enormous quantities of Flash-based content produced through the early 2000s. Almost none of it has been systematically preserved in Korean. The archive problem is global and the volunteer response, for all its real achievement, is patchwork.

What neither project is, and what both are often mistaken for, is a solution. They are a response — one technical, one archival — to a decision that the industry, the platform vendors, and the standards bodies collectively made without factoring preservation into the calculus. The games that came back, came back because a group of people with no job title obliging them to do so decided they should. That is a remarkable thing. It is also, if you think about what an adequate cultural archive actually requires, a damning one.