(Based on analysis of Apache Incubator case studies)
Incubation at the Apache Software Foundation follows many routes. Projects join with different histories, community sizes, and goals. The case studies show that success is not defined by speed but by alignment with ASF values such as openness, transparency, and community-led governance. The following paths summarise what has been observed across multiple projects.
Projects that arrive with diverse contributors, active user adoption, and an ASF-compatible release process often graduate within a year. Incubation mainly serves to verify governance, licensing, and branding alignment.
Typical characteristics
Examples reflected in the data
Projects such as Flink, Iceberg and Arrow followed this route, entering with clear community structures and graduating soon after alignment work was complete.
This is the most common route. Projects start with a small but functional team, expand participation through ASF processes, and graduate once releases, governance, and diversity are stable. Progress is steady rather than fast, balancing learning and autonomy.
Typical characteristics
Initial core team with moderate external engagement
Public development and voting culture established within months
One or two releases per year after early setup
Mentors active but not dominant in decision-making
Examples
Druid, Pulsar and APISIX show this pattern: consistent progress, gradual growth, and sustainable graduation after about two years.
Many projects begin with a small founding team and use incubation to grow broader participation. Progress is measured in new committers, public discussions, and reliable releases rather than speed.
Typical characteristics
Examples reflected in the data
Pinot, Airflow and Kafka showed this pattern with slow early progress followed by strong, sustainable communities.
Large or long-running codebases can require extensive IP clearance, infrastructure migration, or dependency review. These projects may take several years but often become strong ASF communities once established.
Typical characteristics
Examples reflected in the data
NetBeans and IoTDB illustrate this path, where the technical migration was the dominant timeline factor.
Some communities determine that the ASF structure does not align with their preferred workflow or governance model. Withdrawal is a transparent conclusion that leaves both the project and the ASF in a clear position.
Typical characteristics
Examples reflected in the data
Zipkin chose this outcome after maintaining active development outside the ASF structure.
A small number of projects graduate successfully but later slow down or retire. This reflects the normal lifecycle of open-source technology as interests and ecosystems evolve.
Typical characteristics
Examples reflected in the data
Projects such as Mesos, Trafodion, and MXNet followed this trajectory, each contributing valuable lessons in community lifecycle and sustainability.
Some long-running incubations benefited from mentors who stayed engaged throughout, ensuring stability and eventual handoff to independent governance.
Typical characteristics
Examples reflected in the data
Projects such as Superset and ECharts demonstrate how consistent mentorship helped communities mature at their own pace.
Different incubation paths reflect the diversity of open-source projects that join the ASF. Some enter as mature software donations, others as small new communities, and some as large-scale migrations. ASF incubation is designed to accommodate all of these situations by focusing on community readiness and transparency rather than fixed timelines or metrics. The variety of outcomes demonstrates that incubation is flexible by design.
Across more than twenty examples, a few consistent themes appear:
These observations highlight that incubation outcomes depend on continued participation and transparency rather than project age, size, or sponsor.
| Path | Typical Duration | Community Characteristics | Example Outcomes |
|---|---|---|---|
| Established community | Under 1 year | Multi-organisation, regular releases | Flink, Iceberg, Arrow |
| Typical ASF path | 1-2 years | moderate growth, stable governance | Druid, Pulsar, APISIX |
| Emerging community | 1–3 years | Small start, growing diversity | Pinot, Airflow, Kafka |
| Complex migration | 2–5 years | Large codebase, IP clearance focus | NetBeans, IoTDB |
| Withdrawal | Varies | Limited on-list activity | Zipkin |
| Graduation then quiet | 2–3 years | Strong at graduation, later slowdown | Mesos, MXNet, Trafodion |
| Extended mentorship | Several years | Consistent mentor involvement | Superset, ECharts |
Incubation is about building trust, transparency, and shared ownership. Each path adds to the ASF’s collective understanding of how diverse communities grow and sustain themselves in the open.