AWS Builder Center

Stop Calling a Project Open Until the Path to Commit Is Written Down

A README that says "meritocracy" tells you nothing about whether an outsider can actually earn merge rights. This article argues that the defensible measure of openness is whether the contributor ladder is documented, observable, and attainable, because the empirical record shows most projects never write the path down, very few external developers ever cross into the core, and the same role title means different things in different repositories.

Series: developer-tools (8 articles)

  1. …
  2. 8
    Stop Calling a Project Open Until the Path to Commit Is Written Down This article
  1. The word "meritocracy" is older than the evidence behind it
The framing has a clear origin. Fielding's account of the Apache project defines the model directly: the more work a person has done, the more they are allowed to do, and new members are added when a frequent contributor is nominated and approved by the existing voting members [5]. Code changes moved on a minimal-quorum vote, with a small number of positive votes required and a veto reserved for explained objections. That design was built for a specific group of distributed volunteers who could not rely on a manager or a synchronous meeting, and it scaled because granting access was treated as a renewable resource rather than a hoarded one [5].
The principle survives in current foundation practice. The Apache Software Foundation runs a contributor ladder that grows a user through committer and project committee responsibilities, and frames each step as concrete influence over direction rather than a reward [3]. OpenSearch publishes a three-role ladder, Public, Member, and Maintainer, with the responsibilities and privileges of each role stated so anyone can see what the next step requires [4]. These are the projects doing the thing explicitly. They are not the norm.
  1. Most projects never document the path
A study of modern top communities screened the top one thousand GitHub projects down to 734 genuine open-source repositories and read every README, along with linked GOVERNANCE and CONTRIBUTING files, looking for a stated way to gain commit rights. Only 43 projects specified how to apply [1]. That is the headline finding. The other 691 left the boundary between outsider and committer unwritten, which means a contributor cannot know what is expected, and the decision rests entirely on the unstated judgment of the people already inside.
The same research coded the 43 documented policies into a taxonomy of 26 qualifications across nine themes, then ran survival analysis on two rich cases, Node.js with 125 committer transitions and Terraform with 98, to see how the written criteria operated in practice [1]. The practical criteria were broadly consistent with the policies, but the conversion rate was stark: fewer than 2 percent of external developers ever reached committer status, and the probability of gaining rights declined the longer a person had been around rather than rising with tenure [1]. The early period is what matters. Contributors who produced high-quality code, participated in code review, and worked across related projects were the ones who crossed the boundary, while the study judged the community's ability to cultivate people gradually as limited.
![Concept figure showing a five-rung contributor ladder alongside the documented-path and role-drift evidence.]( https://aka.doubaocdn.com/s/S8T8KPGW6m )
Figure by the author (illustrative).
  1. The ladder is not as neutral as it claims
Hold the uncomfortable results rather than polishing them. In the company-backed Terraform project, developers supported by a company entered the trust circle more readily, with a hazard ratio of 1.70, meaning affiliation raised the chance of promotion by roughly 70 percent [1]. That is plausible given the vendor's commercial involvement, but it undercuts the idea that employment is irrelevant. More awkwardly, developers who posted more offensive comments had a slightly higher chance of gaining rights, a hazard ratio of 1.14, which runs against the codes of conduct these same projects publish; the authors note it may reflect a blunt, critical elite style rather than any intended rule [1]. Some written qualifications, such as supporting newcomers, were not meaningfully evaluated at all [1].
A separate analysis of the Eclipse ecosystem reaches a structural version of the same critique. Measuring monthly commits across the hierarchy, the authors conclude the system is not always as fair as it appears and that merit helps only in some cases; meritocracy operates at the lower, project level but gives way to more conventional governance higher up, where board seats are allocated to member organizations rather than earned through commits [6]. Of the 37 projects assigned mentors, only 16 of those mentors had actually committed anything to the project, a reminder that a titled role and a functioning role are different things [6].
  1. Role drift makes titles unreliable
Even when roles are written, the vocabulary cannot be trusted across projects. A cross-project study used Institutional Grammar to decompose every documented role into scope, privileges, obligations, and life-cycle rules across 133 extracted roles [2]. It found role drift: identical titles carry different responsibilities from one repository to another, and different labels describe near-identical work, alongside a maintainer paradox in which a few people accumulate technical, managerial, and community duties at once [2]. A title handed to a contributor on the assumption it means the same thing it meant in the previous project is a misunderstanding waiting to happen. The fix is mechanical. Define each role on those four dimensions rather than borrowing a name, and separate the organizational tier, which sets strategy, from the operational tier that does the daily work, even though maintainers will legitimately bridge both.
  1. A governance contract
Element
Requirement
Failure if skipped
Ladder
Publish every role from user to steering, in order
Outsiders cannot see a path; promotion looks arbitrary
Role definition
State scope, privileges, obligations, and life cycle
Role drift; same title means different work
Entry criteria
Name concrete, observable evidence for each step
Decisions rest on unstated inner-circle judgment
Nomination
State who can nominate and how approval is decided
Sponsorship depends on private access
Inactivity
Define how roles are retired or made emeritus
Defunct members retain rights or lose them abruptly
Evaluation
Audit whether stated criteria are actually measured
Qualifications like mentoring exist only on paper
Distribution
Track whether duties concentrate on a few people
Leadership overload hides behind a flat chart
  1. Rolling it out
Write the ladder before you need to promote someone, and treat the document as living infrastructure reviewed alongside the code rather than a one-time announcement. For each rung, replace adjectives with observable evidence, so "trusted contributor" becomes a specific set of merged changes, reviews, or issue work that anyone can inspect. State the nomination and approval mechanics in the open, including who decides and what quorum is required, so sponsorship is not a private conversation. Add an explicit life-cycle rule for inactive members, distinguishing emeritus recognition from active rights. Then audit the policy against reality: pull the actual promotions and check whether the documented criteria, including the non-code ones, were applied, and watch for duties collapsing onto a small group.


Next steps: check whether your repository states how an outsider earns commit rights, and if it does not, draft a ladder that names observable evidence for every role. Define each role by scope, privileges, obligations, and life cycle rather than by title alone. Add a written nomination and approval process, then review past promotions against the stated criteria and check whether work is concentrating on a few individuals.

References

  1. "How to Gain Commit Rights in Modern Top Open Source Communities?" (734 repositories screened; 43 document a path; under 2% immigrate; survival analysis; hazard ratios reported) —  https://dl.acm.org/doi/pdf/10.1145/3660784 
  1. "Governance in Practice: How Open Source Projects Define and Document Roles," CHASE 2025 (133 roles; role drift; Institutional Grammar; maintainer paradox) —  https://dl.acm.org/doi/full/10.1145/3794860.3794911 
  1. Apache Community Development, "The ASF Contributor Ladder" (foundation guidance; roles from user to maintainer) —  https://community.apache.org/contributor-ladder 
  1. OpenSearch, "Climbing the Ladder: Understanding OpenSearch's Contribution Ladder Policy" (Public, Member, Maintainer roles) —  https://opensearch.org/blog/climbing-the-ladder-understanding-opensearchs-contribution-ladder-policy/ 
  1. Fielding, "Shared Leadership in the Apache Project," Communications of the ACM, 1999 (meritocracy definition; quorum voting; nomination) —  https://roy.gbiv.com/pubs/cacm99\_apache.pdf 
  1. "The Merits of a Meritocracy in Open Source Software Ecosystems" (Eclipse analysis; not always fair; 16 of 37 mentors committed) —  https://dl.acm.org/doi/pdf/10.1145/2642803.2642810 

Series: developer-tools (8 articles)

  1. …
  2. 8
    Stop Calling a Project Open Until the Path to Commit Is Written Down This article
Any opinions in this article are those of the individual author and may not reflect the opinions of AWS.
Enjoyed reading this content? Let the author know!

Your likes, comments, shares, and saves help creators reach more builders.

Loading recommendations

Loading article