Apache Superset

Original Article ↗ Hacker News Discussion ↗

Apache Superset, an open source business intelligence and data visualization tool, is drawing renewed attention as a potential alternative to commercial platforms like Tableau, Power BI, and Looker. Commenters highlight its strengths in SQL-based analytics, dashboarding, embedding, and extensibility, but frequently criticize its learning curve, documentation gaps, and lack of a robust built‑in semantic layer compared with some rivals. Many compare it directly to Metabase, Grafana, Redash, and other OSS tools, weighing Superset’s power and licensing advantages against usability issues and broader skepticism about long‑term investment in Apache‑hosted projects.

Overall Reception

  • Mixed views: some call Superset “phenomenal” and use it in production, including at large companies; others found it unintuitive, buggy, or painful to set up and maintain.
  • Several report successfully replacing Tableau with Superset for cost and openness; others say it has only a fraction of Tableau/Power BI’s features.
  • Many emphasize that it has improved significantly since ~2017–2020, especially with a dedicated company behind it.

Usability & User Experience

  • Multiple commenters say “intuitive” is not accurate; key actions are harder than raw SQL, documentation is sparse, and there are many small UI “papercuts.”
  • Others find the design straightforward, especially the SQL Lab compared to Grafana’s editor.
  • Common complaint: not ideal for non-technical business users; better when engineers or data professionals define datasets and queries.

Comparisons with Other Tools

  • Metabase: Frequently praised as easier to install and more user-friendly, with joins and good self-serve BI; criticized for opinionated UX (models vs questions), limited charts, and paid-only serialization. Many teams choose Metabase for business users and Superset for more customizable or embedded use.
  • Grafana: Seen as observability/time-series focused. Superset better for ad hoc slicing/dicing, non-time-series BI, and SQL-centric workflows.
  • Tableau / Power BI / Looker: Tableau/Power BI viewed as more polished and feature-rich; Superset wins on cost, openness, and extensibility. Lack of a strong built-in semantic layer and join limitations are seen as major gaps vs Looker.
  • Redash / Kibana: Redash seen as slower-moving post-acquisition; Superset a natural successor. Kibana/Elastic better for operational/search use cases, Superset for BI on SQL warehouses.

Features & Architecture

  • Strong points: SQL Lab, rich visualizations (ECharts-based), embedding SDK, row-level security, RBAC, ClickHouse/Druid support, CSS customization, Kubernetes operator.
  • Weak points: joins (especially cross-database), no “thick” semantic layer by default, limited non-time-series flexibility vs commercial tools, no native JSON API source (workarounds exist).

Deployment & Governance

  • Install experience varies: some say Docker makes it trivial; others report dependency issues, upgrade headaches, and data loss in early versions.
  • Discussion around Apache Foundation: some distrust it as a “graveyard” for projects; others note successful, actively maintained Apache projects and value the explicit lifecycle and stewardship.