The editorial argues that every commercial studio attempting to revive Total Annihilation — Cavedog, Gas Powered Games, Uber Entertainment, Petroglyph — either ran out of money or got bought and gutted. Meanwhile, unpaid Spring/Recoil engine volunteers have quietly accumulated 20+ years of pathfinding, projectile, and economy code that no studio has matched, and BAR now has more concurrent players than any commercial TA-lineage release in a decade.
The editorial frames the C++ Spring engine as a living archive of RTS knowledge — two decades of pathfinding tweaks, LOS algorithms, physics-based projectile ballistics, and economy simulation that commercial teams disbanded before they could replicate. Because it's open-source and forkable, a small team of unpaid contributors was able to build a 2026-grade RTS (BAR) on top of it, proving the engine itself is the asset.
By submitting BAR to Hacker News and driving it to 267 points, the submitter elevates a game that has no publisher, no Steam listing, no marketing budget, and a dated homepage — yet sustains 2,000–8,000 concurrent weekend players in 10v10 matches with thousands of units. The implicit argument is that quality and community alone can sustain a competitive RTS at scale.
Beyond All Reason (BAR) climbed to 267 points on Hacker News this week — not bad for a game with no publisher, no Steam page, no marketing budget, and a homepage that looks like it was built in 2011. The pitch is simple: a free, open-source real-time strategy game in the lineage of Cavedog's 1997 *Total Annihilation*, with 10v10 matches, thousands of units on the field, full physics-based projectiles, and a matchmaker that routinely holds 2,000–8,000 concurrent players on weekends.
BAR runs on the Spring engine (now called Recoil), a C++ RTS engine that was itself forked in 2005 from an effort to recreate Total Annihilation in 3D. Spring has been continuously developed for over two decades by a rotating cast of volunteers. BAR is the most polished game built on it — but it's not the only one. Zero-K, another Spring game, has been shipping since 2010 and still has an active ladder.
The codebase is the genre's actual inheritance: 20+ years of pathfinding tweaks, LOS algorithms, projectile ballistics, and economy-simulation code that no commercial studio has matched since Chris Taylor's original team disbanded. You can clone it, you can fork it, and — judging by BAR's quality — you can build a 2026-grade RTS on top of it with a team of unpaid contributors.
The Total Annihilation lineage is a graveyard of commercial failures. Cavedog folded in 1999. Gas Powered Games shipped *Supreme Commander* (2007) and *SupCom 2* (2010), then collapsed into Wargaming. Uber Entertainment Kickstarted *Planetary Annihilation* in 2012, raised $2.2M, shipped a thin product, and effectively abandoned it. Petroglyph's *Grey Goo* (2015) sold poorly. Every studio that tried to make the spiritual successor either ran out of money or got bought and gutted.
Meanwhile, the unpaid volunteers kept the engine alive — and as of 2026, the open-source fork has more concurrent players than any commercial TA-lineage release has had in a decade. This is the kind of outcome that the "open source can't make games" crowd has been confidently wrong about for 15 years. Games-as-software have always been the hard case for OSS: art assets are expensive, balance requires sustained design attention, and players churn if the experience isn't polished. BAR cleared all three bars by accreting contributions slowly over two decades rather than trying to ship a v1.0.
The HN comments are revealing. Multiple commenters who played TA in 1997 say BAR is the first game in the lineage that *feels right* — the projectile physics, the economy curve, the unit-cap-as-strategic-choice. One top comment notes that BAR's lobby has more people in it on a Tuesday afternoon than *Stormgate* (Frost Giant's $40M-funded RTS launched in 2024) has on a Saturday night. That's the gap between a venture-backed studio chasing a genre revival and a community that has been quietly running the revival since 2005.
The lesson for software economics: when the addressable market is too small for VC math but the codebase is permissive and the community is patient, open source can outlast every commercial attempt to serve the same niche. This is the same dynamic that produced Blender (vs. proprietary 3D suites), Godot (vs. mid-tier game engines), and Krita (vs. discontinued raster editors). RTS is just the latest niche where the volunteers outlasted the suits.
There's also a technical story worth flagging. The Spring/Recoil engine handles 10v10 matches with several thousand units per side — physics-simulated projectiles, line-of-sight per unit, terrain deformation — at 30+ fps on midrange hardware. The engine uses a deterministic lockstep model so a 10-player match only syncs commands across the network, not state. That's an architecture most modern engines have abandoned because deterministic floating-point is hard to maintain across CPU vendors. The Spring devs have just… kept maintaining it.
If you're building anything in a niche the big players have written off, BAR is the proof-of-concept that the patient, permissive-license, no-funding path can produce a better product than the funded competition — provided you accept a 20-year timeline. That's not a strategy for a startup, but it is a strategy for a community, and the distinction matters when you're choosing how to license code you might not maintain forever.
For game devs specifically: Spring/Recoil is BSD-licensed and the BAR content is CC-BY-NC. If you've ever wanted to ship an RTS without writing a pathfinder, a unit-balancing system, or a deterministic netcode layer, the work is done. The hard part — finding players — is also partially solved, because BAR has built a lobby and matchmaker you can fork. The genre is not dead; it just stopped being VC-fundable.
For everyone else: this is a useful counter-example next time someone insists open source can't compete on "experience-heavy" products. The argument has always been that games, design tools, and consumer apps need a single coherent vision that committees can't provide. BAR is run by a committee. It has a coherent vision. Both things are true, and the resolution is that the vision is *Total Annihilation, 1997*, and everyone working on the project agrees on what that means.
The interesting question is whether BAR's trajectory generalizes. Stormgate's flameout in 2024–2025 — a studio of ex-Blizzard RTS leads, $40M raised, and a player count BAR beats on a slow day — suggests the genre's commercial revival is over. What replaces it is probably more BARs: small communities forking long-lived engines, shipping slowly, and accepting that their TAM is 10,000 dedicated players rather than 10 million casuals. The same pattern is already visible in MUDs, roguelikes, and immersive sims. RTS just joined the list.
This is amazing to see TA on HN.I grew up playing Total Annihilation in the 90s because my cousin worked for Cavedog and got us a CD for free. It is still one of my favorite games to this day.So many great memories with that game, countless hours playing with my brothers, getting up early to play be
I play Zero-K (a related open source game) a few times a month. I really enjoy their 'cold take' blog series discussing game mechanics and the history of TA derivatives. https://zero-k.info/mediawiki/Cold_Takes
Best RTS ever made, in my subjective opinion, but incredibly toxic player base despite active moderation.I think it's a numbers issue with 16 players in a standard game, and if you play five games in a session, with all chat enabled, you're going to run into a small handful of truly nasty
The game is very mature and well thought RTS. The user experience and performance is beyond anything on the market now. It beats all well known AAA studios attempts at it by far.It's still actively developed and very free to play.Here's a cast of 40 vs 40 players by a former Star Craft 2 p
Top 10 dev stories every morning at 8am UTC. AI-curated. Retro terminal HTML email.
Great game, technically impressive, but the community can be quite mean and toxic. I played for around a year after having played other TA Spring RTS for a few years prior, and if you don’t follow the exact meta of the month (whatever that might be) in terms of build order and things, people can get