Contact

You can reach Reader Signals by email at narang.product@gmail.com.

Reporting an error

We want our numbers to be right. If you think an article contains an error, please email us with:

  • the article title or link,
  • the sentence or number you think is wrong,
  • why you think so, with a link to a source if you have one.

We review each report that includes enough detail to check, against the article’s dataset and sources. If the error is confirmed, we correct the article and add a dated correction note that says what changed. How corrections work is described on the Methodology page.

Other messages

Questions about our data and suggestions for topics are welcome at the same address. Comments are turned off on this site. For questions about how information is handled on this site, see the Privacy Policy.

We do not accept sponsored posts or paid placements.

Popular posts from this blog

opencode and Claude Code: Hacker News stories and npm downloads moved in opposite directions over 30 days

Summary. Between 2026-09-11 and 2026-10-10, only two of the 6 AI coding CLIs tracked here had enough Hacker News stories to compare across windows. For both of them, HN title counts and npm downloads moved in opposite directions: opencode: HN stories changed by -20.0% while npm downloads changed by +19.9%. Claude Code: HN stories changed by +3.9% while npm downloads changed by -27.0%. The two signals are measuring different things. Neither one, alone or combined, measures adoption. The question If HN attention tracked npm downloads, a tool’s share of Hacker News stories would sit close to its share of npm downloads, and the two counts would rise and fall together. The pipeline tested that assumption with a combined metric, the attention-to-download ratio. This is a tool’s share of HN stories divided by its share of npm downloads, both measured within the 6 tracked tools. A value of 1 means the two shares match. Across all tracked tools, the current window contains 239 HN st...
Read more