Analyzing Database Trends Through 1.8 Million Hacker News Headlines

How the analysis was done

I used camelAI with a ClickHouse database of every HN story to do all analysis. You can use it for free with no login here to explore the data interactively yourself.

18 years • 1.8 million headlines • 13 database engines

Hacker News is a real-time barometer of developer excitement. I mined every story title from February 2007 to June 2025 and asked three questions:

• How has headline volume for each database changed over time?
• Which engines are accelerating the fastest right now?
• Which ones are losing steam?

1 · Raw popularity through time

The chart below shows monthly headline counts for the eight most-talked-about engines, plus DuckDB (small base, big growth).

Click on the individual database names to toggle their visibility. Double click to isolate a single database.

PostgreSQL’s curve is a near-monotonic climb; by 2020 it dwarfs every other line. MySQL dominates the pre-2012 era, then flat-lines. MongoDB peaks around 2013, then slips as SQL engines add JSON support. ClickHouse (2016) and DuckDB (2020) rocket up in the analytics renaissance. Redis & SQLite are steady, underscoring their “invisible infrastructure” roles.

2 · Who’s gaining the fastest?

I compared the last 12 months (Jul 2024 – Jun 2025) with the previous 12 months, then fit a linear regression over the last 36 months to capture trajectory.

DuckDB tops the growth chart: half of its lifetime headlines appeared this year. ClickHouse maintains double-digit gains on a larger base, buoyed by vector search & managed services. Supabase benefits from the “Firebase, but OSS” narrative. PostgreSQL’s YoY dip merely reflects an all-time-high 2023; its 3-yr slope (+1.22) is the steepest of all. Cloud-native SaaS engines (DynamoDB, BigQuery, Redshift) are down sharply in share of conversation.

3 · Engines in decline

Headline counts aren’t everything, but they do reveal mind-share:

Why? Maturity – less novelty, fewer “Show HN” launches. OSS competition – Postgres extensions + parquet/iceberg lakehouses replace single-purpose stores. Discussion shift – Cost, lock-in, and serverless topics often omit the product name.

4 · Engagement-Weighted View of Hacker News Database Stories

(July 2024 – June 2025)

Beyond raw headline counts, weighting stories by Hacker News points and comments reveals how intensely developers engage with each database.

Aggregate story count, points, and comments for each database in the last 12 months.

Insights when weighting by points:

• PostgreSQL dominates raw engagement—no surprise given its sheer volume of stories.

• SQLite punches far above its weight:

• Only one-third the story count of Postgres but two-thirds the total points.

• Highest average score per story (40 pts) among mainstream engines.
• Redis consistently sparks high-interest conversations, ranking highly in average points per story.

Insights when weighting by comments:

• High-comment threads often involve operational war stories or novel deployments (e.g., “SQLite in production at the edge”).
• BigQuery and DynamoDB receive respectable per-story points (17 and 7, respectively) yet generate fewer comments, hinting at passive interest rather than debate.
• TimescaleDB and Redshift barely move the needle, confirming their sliding community mind-share.

What the weighted lens adds:

Raw mention counts reveal frequency; weighting by points and comments captures intensity of community reaction.

• SQLite’s high per-story engagement reinforces its status as a beloved workhorse.
• DuckDB garners solid points but fewer comments, typical for emerging tools still new to many readers.
• ClickHouse stories show moderate engagement, suggesting interest but perhaps less day-to-day usage among the broader HN audience.

Taken together, frequency and engagement metrics offer a richer, more nuanced perspective on developer sentiment.

Takeaways

Both raw popularity and weighted engagement metrics show clear trends:

• Open‑source engines drive most new discussion.
• Analytics‑focused stores (DuckDB, ClickHouse) gain as work moves from batch to interactive.
• Slow, steady improvement (PostgreSQL) beats hype cycles.
• Proprietary cloud databases draw fewer submissions over time.

Further analysis

Questions you can run next:

• Which databases appear most in headlines containing "alternative to" or "replacing"?
• What's the ratio of positive (launched, shipped) vs negative (outage, breach) headlines?
• How do vector database mentions (Pinecone, Weaviate, Qdrant) compare to traditional databases?
• Which lesser-known databases are on exponential growth curves?

Run any query on the same dataset here, no login required.