MySQL's New Governance Model: Two steps forward, one step to the side, and one step backwards
After years of MySQL development happening largely behind closed doors, Oracle has been making real progress towards making the MySQL project more open and community-oriented. Thanks and recognition should go to Heather Van Cura and the MySQL team for driving these major initiatives which require heavy lifting on a 30-year-old project. As active members of the MySQL community and as a company committed to open source, VillageSQL is encouraged by Oracle's initiatives towards more open MySQL development.
This week’s announcement of a new governance model and a new Steering Committee for MySQL are positive steps but don't get us all the way to fully open source. This new governance model is the most significant structural change to MySQL's community engagement in a long time. It deserves a genuine read, not reflexive cynicism, and not uncritical celebration either. You can read the announcement from Oracle here and the entire governance model here. AWS posted its commentary here.
To Oracle's credit, the community had been pushing for something like this. In the first part of 2026, hundreds of developers, companies, and community members signed an open letter calling for a neutral non-profit foundation. The OurSQL Foundation was born from that ground-up initiative. Oracle's response to that community groundswell has been top-down in the form of this new governance model, which is generally consistent with how Oracle has handled other open source projects in its portfolio (e.g., Java or OpenOffice).
What's genuinely good here
Any structural change that engages the community and creates defined roles for external organizations and contributors is a major step in the right direction compared to the behind-closed-doors model that preceded it. More importantly, Oracle's layoffs left perceived gaps in MySQL's development capacity. Getting AWS and Google to commit engineering resources to MySQL directly is a bet that hyperscaler developers can help fill those gaps and potentially accelerate development in ways Oracle's team couldn't alone. AWS has already outlined its plans to contribute. That's real engineering capacity entering the codebase.
The PostgreSQL parallel is worth naming here. Today, both Google and AWS have multiple engineers contributing directly to PostgreSQL mainline development. AWS employs PostgreSQL committers, earned through years of contribution, not purchased through a seat. Google contributes to logical replication and major version infrastructure. Their investments benefit the whole community, even if they're also motivated by competitive advantage. MySQL getting the hyperscalers more engaged is a genuine win for the ecosystem.
What "open governance" actually means
The governance model that Oracle published uses the vocabulary of open source governance correctly in some places and awkwardly in others. The contributor ladder, Contributors → Committers → Project Leads → Core Project Leads → Steering Committee, is a real structure with real criteria. For example, 5 to 10 accepted contributions to achieve Committer status; 30 to 50 contributions in a subsystem to achieve Project Lead.
The concern is who controls advancement at every rung. Committers are currently Oracle employees and nominated by Oracle. Project Leads are nominated by Oracle initially. The Steering Committee's non-Oracle seats were chosen by Oracle for a 2-year initial term, after which community elections become possible. The Core Project Lead role, the one that owns release and cross-component decisions, has no stated path for any external person to hold it.
Fedora (a Red Hat/IBM-owned project) is an interesting counterpoint. It is primarily maintained by Red Hat employees, and Red Hat hires the Fedora Project Leader. But the Fedora Engineering Steering Committee, the body that governs technical policy, is fully community-elected. A developer at any company can stand for election, win, and hold a top-level governance role without Red Hat's blessing at any step. That's what a contributor ladder with real community teeth looks like.
PostgreSQL has no corporate owner. Committers are selected through an annual peer process: existing committers nominate candidates on a public mailing list, deliberate openly, and the core team extends invitations based on sustained technical merit. 463 people contributed to PostgreSQL 17. No one buys a seat.
While Oracle's model uses the language of open source, it more closely resembles a foundation governing board model (where hyperscalers and other companies purchase membership tiers and receive board representation). Foundation boards operate this way for good reasons; hyperscalers and other companies fund operations, and having skin in the game is appropriate for a funding body. But an open source technical steering committee probably shouldn't work the same way. AWS and Google received Steering Committee seats without going through the contributor ladder at all. That's not a ladder; that's a pay-to-play model.
Who didn't get a seat
We applaud Oracle for its openness and pre-announcements around moving to a GitHub-based bug tracker and development cycle, its hosting of community and contributor events, and its willingness to accept community-authored patches and features.
This new governance model and Steering Committee represent a reversion to the historical model for Oracle, though. The existence of the Steering Committee, its structure and rules, and selection criteria for initial external seats was never publicly announced or explained in advance.
We are optimistic this was done this way to bootstrap the governance model and get the hyperscalers bought in, but we'll have to wait a full 2 years to know for sure. The absence of active independent MySQL community members or the community-led foundation from the initial composition of the Steering Committee is notable and worrisome. While this is a step in the right direction, it doesn't go all the way to addressing the concerns of the MySQL community.
The structural problem
Good open source governance does two things for outside contributors: it gives them an incentive to contribute, and it gives them a path to leadership. The incentive is recognition: committer status, project lead authority, influence over direction. The path is merit-based and doesn't require vendor approval at every step.
This new MySQL model doesn't fully provide either. In practice, Oracle appears to be outsourcing significant development to hyperscaler contributors. The problem is Oracle (and now the hyperscalers) dictate what gets in and what stays out. Oracle nominates advancement at every level. The Core Project Lead role has no external path. Strategic authority stays with Oracle, not the community.
That's not necessarily fatal. PostgreSQL thrives with a pure meritocracy; Linux thrives with Linus as a benevolent dictator; Red Hat runs Fedora and Fedora is widely respected. The governance model arguably matters less than whether the project ships good software. But announcing a structure as "open governance" when the primary vendor and other large corporations (the hyperscalers) control the top of every ladder invites a level of scrutiny and skepticism that wouldn't otherwise exist.
Where we stand
We're genuinely optimistic that MySQL's new direction will result in more work getting done and more benefit for the users and companies that deploy MySQL. Oracle's layoffs created real perceived capacity concerns; hyperscaler engineering fills some of them; community contributions will fill others. Movement toward any kind of structured external participation is better than the closed-door model that preceded it. The seat allocation question, specifically the absence of independent MySQL companies from the initial Steering Committee, is something we'll continue watching as this all evolves.
This blog post is an initial reaction. We hope our initial concerns are proven wrong and this is the first step toward a truly open MySQL project with direct input from the community.
