• Created by Unknown User (jmclean) on Nov 30, 2025 3 minute read

You are viewing an old version of this page. View the current version.

Compare with Current View Page History

Version 1 Current »

Incubator by the Numbers (2015–2025) - A concise, data-driven view of how incubation has shifted over a decade

This brief summarises ten years of Incubator data (2015–2025) covering podling lifecycles, releases, community growth, governance activity, and system behaviour. The aim is to highlight observable trends using consistent, decade-long signals.

1. A Smaller, More Stable Incubator

  • 2015–2019: typically 45–63 active podlings
  • 2023–2025: typically 28–32 active podlings

Intake has declined, and long-running stalled podlings from the mid-2010s have largely been resolved. A smaller Incubator naturally produces fewer graduations, fewer release votes, fewer governance threads, and fewer problem cases.

2. Outcomes: Faster, Not Easier

Graduation timing

Graduation timing tracks podling population:

  • High-podling years: 20–42 months median
  • Low-podling years: 13–18 months median

Modern podlings progress through early milestones sooner, due to both:

  1. Structural effects: fewer podlings → fewer bottlenecks
  2. Behavioural effects: earlier first releases → earlier community and governance growth

Graduation vs retirement

Fully matured cohorts (2015–2020) show stable outcomes:

  • Graduation: ~70–80%
  • Retirement: ~20–30%

Retirement typically occurs 40–60 months after entry. Cohorts entering after 2019 have not yet reached this window, so apparent declines in retirements are age-based, not behavioural.

3. Releases: Earlier, More Often, More Reliable

Time to first release

  • 2015–2018: ~6–7 months median
  • 2019–2024: ~4–5 months median

Release attempts now occur earlier and more consistently.

Release volume

Normalised release votes:

  • 2015–2022: ~4.0–4.5 per podling
  • 2023–2024: ~6.1–6.3 per podling

Technical issue decline

Technical issues (LICENSE/NOTICE/DISCLAIMER, naming, packaging) are significantly less common than in the mid-2010s. Most modern release issues are behavioural rather than technical.

4. The First Release Effect

Across all cohorts, the timing of the first release is the strongest early indicator of long-term trajectory.

Podlings that release sooner tend to:

  • add governance members earlier
  • make recurring releases
  • show stable dev@ activity
  • attract contributors beyond the initial team
  • graduate far more reliably

Podlings that substantially delay their first release show the same early-stall patterns seen in most historical retirees.

5. Timing of Early Milestones

Graduating podlings consistently reach key milestones earlier:

  • first release: ~4–7 months
  • first community addition: ~6–13 months

Retiring podlings show opposite patterns: late or absent releases, no governance additions, long quiet periods, and work concentrated in 1–2 individuals.

6. Community Growth: First Governance Addition

Using the earliest of the first committer or first PPMC addition (excluding bootstrap), modern podlings add their first new governance participant significantly sooner than mid-2010s cohorts.

  • 2015–2018: typically ~10–11 months
  • 2019–2024: typically ~5–6 months

Across the decade, first governance additions occur earlier, with less variance and far fewer long delays. This aligns with earlier releases, more consistent progression, and fewer stalled podlings.

7. Governance on dev@

Successful podlings show regular dev@ governance activity and fewer extended quiet phases than in the mid-2010s. Normalised dev@ governance load remains stable across the decade and aligns with expectations for project-level decision-making.

8. Governance Load on incubator-general

Governance threads per podling

  • generally 7–12 per podling per year, stable throughout the decade

Release votes per podling

  • 2015–2022: ~4.0–4.5
  • 2023–2024: ~6.1–6.3

Shift in activity mix

Vote threads form a slightly larger share of general@ activity in recent years. Non-vote governance threads per podling are slightly lower in later years, while vote-related threads have increased. Oversight per podling remains stable once podling count is normalised.

9. Cohort Behaviour

Older cohorts (2015–2017)

Show full life cycles: early graduates and late retirees.

Newer cohorts (2018–2024)

Show early graduates, but few/no retirements due solely to cohort age. Only time-to-graduation has changed significantly; historically, retirement patterns remain consistent.

10. Early Indicators of Podling Health (Descriptive Only)

Common signals among graduates

  • first release within 4–7 months
  • governance additions within ~1 year (>60 days)
  • recurring release activity
  • visible dev@ participation
  • multiple active contributors

Common signals among retirees

  • delayed/absent first release
  • no governance additions
  • no new committers after the first year
  • extended quiet periods
  • work dominated by 1–2 individuals

11. Summary: What Has Actually Changed

Structural

  • Active podlings dropped from 45–63 to 28–32
  • Intake is lower
  • System-wide activity tracks podling count

Behavioural

  • earlier first releases
  • earlier community additions
  • more frequent release attempts
  • fewer technical issues in releases
  • fewer long-running stalled podlings

Cohort-age effects

  • apparent declines in retirements are due to age, not standards
  • mature cohorts remain ~70–80% graduates, 20–30% retirees

Oversight

  • dev@ and general@ governance per podling: stable
  • release-vote load per podling: increased
  • oversight remains consistent after normalisation

Taken together, the data shows a program that is smaller, more consistent, and more predictable than a decade ago, operating with stable oversight in a context of clearer expectations and earlier podling progress.

  • No labels