Skip to content

Blog: From Cron to Durable Execution — A Primer on Workflow Engines#1350

Open
rubenfiszel wants to merge 12 commits intomainfrom
blog/workflow-engines-primer
Open

Blog: From Cron to Durable Execution — A Primer on Workflow Engines#1350
rubenfiszel wants to merge 12 commits intomainfrom
blog/workflow-engines-primer

Conversation

@rubenfiszel
Copy link
Copy Markdown
Contributor

Summary

Long-form technical blog post comparing workflow engines, targeting HN audience.

Engines covered in depth: Airflow, Prefect, Temporal, Inngest, Windmill WAC, Windmill Flows
Honorable mentions: Restate, DBOS, Hatchet

Content highlights

  • Theory section: persistence spectrum, state representation trade-offs, worker architecture
  • Same 3-step workflow implemented in all 5 engines side-by-side
  • Dynamic control flow comparison (try/catch vs DAG branches)
  • Real benchmark numbers (Windmill v1.683 vs Temporal v1.29)
  • GitHub stars landscape chart (8 engines)
  • Honest trade-off matrix

SEO keywords

workflow engine, durable execution, Airflow, Temporal, Prefect, Inngest, Windmill, workflow-as-code, DAG scheduler, event sourcing

Notes

  • Benchmark data from windmill-labs/windmill branch wmap-benchmark-metrics
  • No cover image yet — needs one before publishing
  • Numbers may need updating when run on standardized CI hardware

🤖 Generated with Claude Code

…ct, Temporal, Inngest, Windmill)

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
@cloudflare-workers-and-pages
Copy link
Copy Markdown

cloudflare-workers-and-pages bot commented Apr 15, 2026

Deploying windmilldocs with  Cloudflare Pages  Cloudflare Pages

Latest commit: 35e6e87
Status: ✅  Deploy successful!
Preview URL: https://8438aeca.windmilldocs.pages.dev
Branch Preview URL: https://blog-workflow-engines-primer.windmilldocs.pages.dev

View logs

rubenfiszel and others added 11 commits April 15, 2026 11:33
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
…a are by convention

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
…ugh DB is universal

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
…ly, no visual editor

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
…rt is high

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
… not language

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- Reframe "Three Generations" as two generations plus a visual UX over
  Gen 1 (flow builders are an authoring mode, not a distinct generation)
- Wrap the 5-engine code comparison and dynamic-control-flow comparison
  in Tabs so readers don't have to scroll top-to-bottom
- Add a Benchmark Methodology section explaining why only Temporal is
  compared and defining seq_N / par_N / fan_out_N; note repro repo is
  planned
- Replace most ASCII box/sequence diagrams with React components:
  DAG scheduler model, Airflow state machine, Temporal replay sequence,
  Inngest HTTP roundtrips, persistence spectrum, worker architecture,
  GitHub stars bar chart
- Reduce em dashes across the post (periods/colons/commas where they
  were lazy sentence joiners; kept genuine parentheticals)

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Dagster's ~13.5K stars put it above Hatchet/Inngest/Restate/DBOS, and
its software-defined-asset model is a genuinely different authoring
paradigm on top of the Gen-1 execution model — worth a short entry
even though the underlying runtime is similar to Airflow/Prefect.

Also update the description, intro, section title, and GitHub stars
chart to include Dagster.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant