Video.js creator reclaims project from PE, ships 88% smaller rewrite

Video.js — the open-source video player embedded on Amazon, LinkedIn, Dropbox, and enough other sites to reach billions of monthly users — just got its biggest update in 16 years. And the backstory matters as much as the code.

Here's what happened: private equity acquired the company behind Video.js and, in classic PE fashion, fired the maintainers. The project limped along with a skeleton crew doing their best against a dated architecture. The original creator decided to take it back, recruiting Sam from Plyr, Rahim from Vidstack, and other collaborators to do a ground-up rewrite.

The result is Video.js v10 beta, and the headline number is stark: 88% smaller. That's not a typo, and it's not comparing minified-to-unminified. The rewrite shed nearly nine-tenths of the bundle weight.

For frontend developers, that number alone justifies attention. Video players are notoriously heavy dependencies — they sit on critical rendering paths, they block interaction on mobile, and they're one of those libraries teams include once and never revisit. An 88% reduction in a dependency used at that scale isn't an optimization; it's a rearchitecture.

The collaboration model here is worth noting too. Rather than forking or competing, creators of three separate video player projects (Video.js, Plyr, Vidstack) pooled their efforts. In an ecosystem where fragmentation is the default, this kind of consolidation is rare and welcome. Each of these developers has shipped production video infrastructure independently; combining that experience into a single project raises the floor for the entire category.

The broader lesson is one the open-source community keeps relearning: PE ownership and open-source stewardship are structurally misaligned. PE optimizes for extraction — reducing headcount, cutting costs, milking existing revenue. Open source requires the opposite: sustained investment in maintenance, community responsiveness, and architectural evolution. When the incentives diverge, the project rots. Video.js is just the latest example, following similar patterns at projects like Elasticsearch (the Amazon fork), Redis (the licensing pivot), and countless smaller libraries that quietly decay under absentee ownership.

What makes this story different is the recovery. Most PE-gutted open-source projects don't get reclaimed by their original creators with a coalition of talented contributors. Most just slowly die while Stack Overflow answers pointing to them accumulate.

If you're currently shipping Video.js in production: the v10 beta is worth evaluating now, even if you wait for stable. An 88% bundle reduction with maintained API compatibility (if they deliver on that) is the kind of upgrade that pays for itself in Core Web Vitals alone.

If you're not using Video.js: this is still a case study in what happens when critical infrastructure gets the investment it deserves after years of neglect. The 88% number tells you how much dead weight accumulates when no one's empowered to make breaking changes.