This summary describes the end-to-end process for generating the Incubator Field Guides from Apache Incubator mailing list discussions. The intent was to capture real, observable governance patterns.

Purpose

The purpose of this work is to make Incubator governance experience accessible and reviewable.

The Field Guides are intended to:

• help people new to the Incubator (mentors, PPMC members, IPMC members, and proposers) gain a realistic understanding of how the Incubator typically operates,
• surface the kinds of issues that recur in practice,
• and provide that context without requiring readers to work through years of mailing list archives.

Taken together, the guides offer a practical view of how the Incubator has operated over the last five years, based entirely on public discussions.

1. Identify Candidate Threads

Emails were grouped into threads using normalised subjects (reply prefixes removed). Threads were then selected using two threshold regimes based on:

• total messages per subject-thread
• unique senders per subject-thread

Threads selected:

• Tier 1 : ~30+ messages and ~ 12+ unique senders
• Tier 2 : ~18+ messages or 9+ unique senders

Thread counts (from the full scan)

• Total threads scanned: 2940
• Tier 1 threads selected: 39
• Tier 2 threads selected: 516

This approach intentionally prioritises high-volume threads and may miss low-volume threads that are nevertheless important.

2. Group Selected Threads by Subject Area

After candidate threads were identified, the selected threads were grouped into subject areas such as:

• proposals and incubation readiness
• governance structure and roles
• licensing, provenance, and IP clearance
• naming, branding, and external identity
• releases and distribution compliance

3. Consolidate Common Threads

Many selected threads covered the same underlying issue or raised the same questions in slightly different ways. Where this occurred, related threads were consolidated.

4. Extract and Review Emails by Group

For each subject-area group, emails were extracted and reviewed end-to-end. During review, many threads were found to be low-signal for governance learning, such as:

• procedural votes with no substantive discussion
• routine announcements and administrative confirmations
• tracker notifications or operational noise
• threads unrelated to governance mechanics

5. Select Governance-Relevant Content Only

Within each group, only content that appeared explicitly in the emails was retained, including:

• questions raised by reviewers
• objections that blocked progress until they were resolved
• requests that exposed governance mechanics
• dicussions that led to a resolution

6. Anonymise and Generalise Examples

Where examples were used, they were:

• anonymised (projects and individuals not named)
• generalised to focus on the governance pattern rather than a specific dispute

The intent was to preserve institutional learning without creating project-specific case studies.

Outcome

The result is a set of Field Guides that:

• are grounded entirely in public Incubator discussions,
• reflect how the Incubator has operated in practice over recent years,
• make governance experience easier to absorb for newcomers,
• and preserve institutional knowledge without requiring readers to review years of mailing list archives.