4 ms·
https://jaytaylor.github.io/hn-live2 https://jaytaylor.github.io/hn-live2 is doing it though.
by smusamashah 1y ago
https://jaytaylor.github.io/hn-live2 https://jaytaylor.github.io/hn-live2 is doing it though.
- zamadatix 1y agoAs does hnstream.com from the sourced sample comment itself. Both just traverse the parent id until it's the root (article). It takes more queries, but the API is not rate limited.
- krapp 1y agoIt wouldn't take more queries if the comments were cached. It could probably be done entirely in memory, HN's entire corpus can't be that large. If one were to start at the page endpoints (eg /topstories) one could add references to origin ids while preloading comments, and probably cover the most likely to be referenced ID, and even make traversal up the tree even more efficient.