Flashpoint

A preservation project that collected the games themselves at scale, together with the environment needed to launch them.

The end of Flash did not arrive without warning. Adobe announced the retirement years in advance, a deadline that gave a small number of people enough time to do something deliberate about what would be lost. Most of what they built relied on emulating the runtime — replicating the software environment that games ran inside, so that individual titles could be played without needing the original plugin. BlueMaxima's Flashpoint took a different approach: it went after the games themselves, in bulk, and brought the launcher with them.

Collecting at scale

Flashpoint began in 2018, in the period between Adobe's announcement and the actual shutdown, when it was still possible to retrieve games from servers that had not yet gone dark. The project was started by an individual going by the handle BlueMaxima and grew into a community effort. The core idea was straightforward in conception, extraordinarily labour-intensive in execution: download every web game you can find, capture the assets and the metadata, and bundle everything with a proxy system that would redirect the game's network calls to locally stored files rather than the internet.

That proxy approach matters. Many browser games were not self-contained objects. They called out to external servers for assets, for score data, for licence verification, for next levels. A raw download of the SWF or Shockwave file was not enough to make the game run correctly years later; the network layer had to be faked. Flashpoint's solution was to route traffic through a local proxy that intercepted outbound requests and served locally stored responses, so the game believed it was online when it was not. This is preservation by deception, in the most respectful sense of the phrase — the software runs in the environment it expects.

The collection grew rapidly. By the time Flash support officially ended on 31 December 2020, Flashpoint already held tens of thousands of titles. The project continued expanding after that date, working from archived copies, community donations and software recovered through the Internet Archive. At various points in its public release history, the curated collection — Flashpoint Ultimate, the full local installation — ran to hundreds of gigabytes of content. A smaller, streaming-capable version called Flashpoint Infinity allowed selective downloading of individual titles rather than the whole archive, which made the project accessible to people without the storage for the complete set.

What it actually preserves

Flashpoint is not limited to Flash. That name is an artifact of its origins; the project extended its scope to cover Shockwave titles, Java applets, games built on the Unity Web Player plugin, HTML5 games, and content from several other now-defunct runtimes. This breadth is a direct consequence of the underlying logic: the threat to web games was never specific to one format. It was the general dependency on plugin architectures that browsers were phasing out, on servers that would eventually be switched off, on a business model that assumed content was ephemeral. Flashpoint treats that whole ecosystem as its subject.

The metadata work is as important as the binaries. Each entry in the Flashpoint database carries game title, developer, publisher, platform, release year where known, a description, and tags. This is cataloguing work — the kind that libraries do, and that is usually absent from informal preservation efforts. Without it, a collection of raw files is an undifferentiated mass. The structured data is what makes the archive navigable and, eventually, citable.

The project also made deliberate decisions about what to exclude. Content that had surviving commercial availability was generally passed over; the focus was on software that had no other accessible home. The project's curatorial choices were not always consistent — a collection this size, built by volunteers across years, will have edges — but the intent was to concentrate effort where loss was otherwise certain.

The runtime problem and how Flashpoint addresses it

There is a genuine intellectual tension between the Flashpoint model and the emulation-first approach taken by institutions like the Internet Archive. Emulating the runtime, rather than the game, is the more elegant preservationist position: if you can faithfully recreate the Flash Player, every SWF ever made is theoretically playable, without having to track down and store each file individually. The Ruffle project, an open-source reimplementation of the Flash runtime written in Rust and designed to run inside a modern browser, pursues exactly this strategy.

Flashpoint's approach is complementary rather than competing. Ruffle cannot yet handle every ActionScript 3 behaviour faithfully, and there are categories of content — particularly Shockwave files and Java applets — where no general emulator has achieved comprehensive compatibility. Emulation of a runtime is a moving target: it gets better over time, but it is never finished. In the meantime, Flashpoint provides a working environment today, not a promissory note for future compatibility.

The project ships the actual runtime executables — the real Flash projector, the real Shockwave player — alongside the games, which means it depends on those binaries continuing to function on current operating systems. That is a different kind of fragility from the emulation problem, not an absence of it. As operating systems evolve, standalone projector binaries may eventually stop running on native hardware, at which point the project would need to shift toward running its bundled runtimes inside emulated operating system environments. Some preservation thinking about Flashpoint already anticipates this transition.

Scale as a preservation argument

What Flashpoint demonstrated, more than any technical innovation, is that the scale problem in web game preservation is tractable. The platform had accumulated hundreds of thousands of distinct titles across its history, created by individuals, small studios and large publishers alike, distributed freely and often without documentation. That made it easy to assume the task was hopeless. Flashpoint is evidence that organised, motivated communities can cover enormous amounts of ground when they have a clear method and share their work openly.

The Video Game History Foundation has documented, in the context of console and PC software, how large a proportion of games from previous decades have no known surviving copy. The situation for web games is different in character — the distribution method was the internet, so at least some copies typically survived in crawls and caches — but the risk of gradual attrition was real. Flashpoint reduced that risk significantly, for an entire era of a platform, in a window of a few years. That is the achievement.

BlueMaxima's Flashpoint took a different approach: it went after the games themselves, in bulk, and brought the launcher with them.