5 ms·

Related, I am learning to code and in the process I built this simple website that randomly shows the most upvoted HN comments: http://www.opusnota.com/hnbc ht

by soneca 9y ago

Related, I am learning to code and in the process I built this simple website that randomly shows the most upvoted HN comments: http://www.opusnota.com/hnbc http://www.opusnota.com/hnbc

jmstfv 9y ago
How do you fetch comment scores?
minimaxir 9y ago
HN comment scores are not public, so I am unsure how you find the "most upvoted comments"
- RodericDay 9y ago
  top of every story probably
  - minimaxir 9y ago
    That is a very bad metric because comment rankings are a function of score and time.
- soneca 9y ago
  Algolia's API show votes for older comments (not sure how old is the line when it stops showing). Edit: naturally, there is a strong bias for older comments in the list. Code I use: <script> $(document).ready(function() { var lasthours = ""; $("#refresh").on("click", function() { var html = ""; var url = "https://hn.algolia.com/api/v1/search?tags=comment&numericFilters=points>120&page="; var page = Math.floor(Math.random() * (20)) + 1; url += page.toString(); var position = 0; position = Math.floor(Math.random() * 20); $.getJSON(url, function(json) { var hits = json.hits; var html = "<p>"+hits[position].comment_text+"</p><p>Story: <a target='_blank' href='"+hits[position].story_url+"'>"+hits[position].story_title+"</a><br><a target='_blank' href='https://news.ycombinator.com/item?id="+hits[position].story_id+"'>comment by "+hits[position].author+"</a></p>"; $("#quote-text").html(html); }); }); }); </script>
  - minimaxir 9y ago
    The Algolia API cutoff for comment scores is October 2014. (which I know because it was cut off after I made a blog post on the subject: http://minimaxir.com/2014/10/hn-comments-about-comments/ http://minimaxir.com/2014/10/hn-comments-about-comments/) There have been quite a few good comments since then. It is worse to make an incredibly biased list than no list at all.
    - soneca 9y ago
      It is just a learning project. Feel free to not access it and downvote my comment advertising it if you think it is dangerous.
      - minimaxir 9y ago
        My comment was not a criticism, but a warning. It is very important to note where the data is coming from and any associated caveats which may influence the results.
- TeMPOraL 9y ago
  https://news.ycombinator.com/bestcomments https://news.ycombinator.com/bestcomments?
  - r3bl 9y ago
    Been using HN regularly for about two years now, and have just figured out two of the features from this topic: /bestcomments and /favorites?id=<USERNAME>.
    - TeMPOraL 9y ago
      You might want to check out "Lists" link at the bottom of the page; there's a bunch of things there.

by soneca 9y ago

Related, I am learning to code and in the process I built this simple website that randomly shows the most upvoted HN comments: http://www.opusnota.com/hnbc http://www.opusnota.com/hnbc

jmstfv 9y ago

How do you fetch comment scores?

minimaxir 9y ago

HN comment scores are not public, so I am unsure how you find the "most upvoted comments"

RodericDay 9y ago
top of every story probably
- minimaxir 9y ago
  That is a very bad metric because comment rankings are a function of score and time.
soneca 9y ago
Algolia's API show votes for older comments (not sure how old is the line when it stops showing). Edit: naturally, there is a strong bias for older comments in the list. Code I use: <script> $(document).ready(function() { var lasthours = ""; $("#refresh").on("click", function() { var html = ""; var url = "https://hn.algolia.com/api/v1/search?tags=comment&numericFilters=points>120&page="; var page = Math.floor(Math.random() * (20)) + 1; url += page.toString(); var position = 0; position = Math.floor(Math.random() * 20); $.getJSON(url, function(json) { var hits = json.hits; var html = "<p>"+hits[position].comment_text+"</p><p>Story: <a target='_blank' href='"+hits[position].story_url+"'>"+hits[position].story_title+"</a><br><a target='_blank' href='https://news.ycombinator.com/item?id="+hits[position].story_id+"'>comment by "+hits[position].author+"</a></p>"; $("#quote-text").html(html); }); }); }); </script>
- minimaxir 9y ago
  The Algolia API cutoff for comment scores is October 2014. (which I know because it was cut off after I made a blog post on the subject: http://minimaxir.com/2014/10/hn-comments-about-comments/ http://minimaxir.com/2014/10/hn-comments-about-comments/) There have been quite a few good comments since then. It is worse to make an incredibly biased list than no list at all.
  - soneca 9y ago
    It is just a learning project. Feel free to not access it and downvote my comment advertising it if you think it is dangerous.
    - minimaxir 9y ago
      My comment was not a criticism, but a warning. It is very important to note where the data is coming from and any associated caveats which may influence the results.
TeMPOraL 9y ago
https://news.ycombinator.com/bestcomments https://news.ycombinator.com/bestcomments?
- r3bl 9y ago
  Been using HN regularly for about two years now, and have just figured out two of the features from this topic: /bestcomments and /favorites?id=<USERNAME>.
  - TeMPOraL 9y ago
    You might want to check out "Lists" link at the bottom of the page; there's a bunch of things there.