HOW IT WORKS

A video catalog shaped by Hacker News recommendations

SignalTube starts with public Hacker News comments that link to YouTube videos, then organizes those references into searchable topics and recommendation context.

1. Find a video in a comment

The source dataset records videos linked in Hacker News comments from 2007 to 2026. This site uses aggregate stance labels and source links; it does not republish full comment text.

2. Read the recommendation context

Each video shows aggregate stance counts and short notes about how the linked comment used the video in its thread. Comment bodies are not republished; each note links to its original Hacker News comment. Strong recommendations are separated from contextual references, jokes, and critical mentions.

3. Browse or tune your feed

Use topic pages, content-type filters, and HN recommendation history to explore. Likes, “not interested,” Watch later, history, and topic preferences adjust a feed locally in your browser. You can also import a Google Takeout watch-history file without uploading the file.

How “best” and “hidden gems” work

Best sorts by the source dataset’s recommendation score, which weights distinct months and strong recommendations. Hidden gems here means videos with strong HN recommendation labels and fewer distinct recommenders. The supplied export does not include YouTube view counts, so this page does not claim a view-count threshold.

Data source and limitations

HackerTube’s October 31, 2026 export is licensed under ODC-By 1.0 and credited on the About page. Its summaries and classifications are generated from dataset fields and may be wrong. Browse the topic archive or the dataset guide.