Neat! I wonder what happened here:<p><a href="https://hacker-news-analytics.christianfei.com/stats/Show%20HN%3A%20A%20beetle%20generator%20made%20by%20machine-learning%20zoological%20illustrations" rel="nofollow">https://hacker-news-analytics.christianfei.com/stats/Show%20...</a><p>> #6 (0) 184 (1) 19 (0) Show HN: A beetle generator made by machine-learning zoological illustrations 2020-01-02T17:08:54.566Z<p>> #6 (-1) 183 (149) 19 (0) Show HN: A beetle generator made by machine-learning zoological illustrations 2020-01-02T17:06:55.781Z<p>> #7 (3) 34 (-148) 19 (0) Show HN: A beetle generator made by machine-learning zoological illustrations 2020-01-02T17:04:56.732Z<p>> #4 (0) 182 (1) 19 (0) Show HN: A beetle generator made by machine-learning zoological illustrations 2020-01-02T17:00:54.504Z<p>Also, I find myself wishing for a "new comments since your last visit" like Reddit Gold gives you, or some way to "subscribe" to a node of comments and its replies. Anyone have any good solutions for this?
Weird to see posts like this that randomly spike in the number of votes<p><a href="https://hacker-news-analytics.christianfei.com/stats/Exploiting%20Wi-Fi%20Stack%20on%20Tesla%20Model%20S" rel="nofollow">https://hacker-news-analytics.christianfei.com/stats/Exploit...</a>
I think it should identify posts by ID rather than by post title - it seems that reposts and dupes confuse it (for example, the current top post was posted a total of three times per HN search: <a href="https://hn.algolia.com/?dateRange=all&page=0&prefix=true&query=The%20polygons%20of%20another%20world&sort=byPopularity&type=story" rel="nofollow">https://hn.algolia.com/?dateRange=all&page=0&prefix=true&que...</a>, which causes the apparent history in this tool to bounce around a lot: <a href="https://hacker-news-analytics.christianfei.com/stats/The%20Polygons%20of%20Another%20World" rel="nofollow">https://hacker-news-analytics.christianfei.com/stats/The%20P...</a>)
Slightly related, but has anyone made an automated fuzzy, dupe detector for posts or topics? I keep thinking it would be nice to browse HN using an explicit graph constructed with some NLP clustering, instead of an implicit graph created by links in the comments.
I've frequently wanted an 'evocative' score that would provide a metric for the ratio of up vs. down votes among the comments.<p>What's more contentious, 'Tabs vs. Spaces', or, 'No Go, Python'.
<a href="http://hnrankings.info/" rel="nofollow">http://hnrankings.info/</a> does a pretty good job about this as well, though its individual post viewing capabilities are limited.
Interesting, is there a reason you are sending the data to the client so regularly? Is that being triggered every time something changes?<p>Also how are you liking SSE? I've played with WS but never SSE.