Hacker News new | past | comments | ask | show | jobs | submit login
Ask HN: Does dang read every HN comment?
44 points by johnwheeler 15 days ago | hide | past | web | favorite | 7 comments
Can you give us an idea of what percentage of content on HN you typically consume in a day? Is it 100%, 50%?



Based on his admiration of Metafilter’s moderation and how much has been written in Metatalk about how their moderation works I’d say he spends a lot of time responding to flags and emailed notes, watches new and some special lists/views, and maybe keeps open the few threads he intuits to be contentious (possibly with access to a linear view of comments being added to an article, and extra metadata appearing?)

Good, efficient moderation requires internal tooling and view development.


https://news.ycombinator.com/user?id=sctb is also a moderator and there is a undisclosed number of hidden semi-moderator, and some automatic filters and scripts, and enough flags from users can kill a bad story. So I guess 20%.

If you see something too bad or too good, you can send an email to the mods: hn@ycombinator.com They usually reply soon, but it is a manual process, so use it wisely.


The recent post [1] "The Loney Work of Moderating Hacker News", should shed some light on how he moderates.

If you go to the [2] Algolia API, and filter "Comments in the last 24 hours". There's around 8,000 results. It's not really possible to keep track of that many comments and conversations (if he is a human that is).

My guess is that the moderation tools help him monitor for content that needs urgent attention, to keep the community and discussion healthy... so my final answer ~10% (1,000 comments per day)

[1] https://news.ycombinator.com/item?id=20643052

[2] https://hn.algolia.com/?query=&sort=byPopularity&prefix&page...


Not physically impossible, 24h a day thousands of comments s day. I imagine grep or similar is used to find bad comments.


Depends if he consumes it through the website or has a reader or stream-like view. I'm imagining a twitter like feed that could elevate controversial comments to his immediate attention.


Moderation will always need the human touch, but it seems to be ripe for some sort of ai or graphing tool to help identify where it's likely to be needed.

All thanks to dang and sctb anyway. :)


Yes




Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact

Search: