3 ms·

https://news.ycombinator.com/item?id=36680674 https://news.ycombinator.com/item?id=36680674 https://news.ycombinator.com/item?id=36683756 https://news.ycombina

https://news.ycombinator.com/item?id=36680674 https://news.ycombinator.com/item?id=36680674 https://news.ycombinator.com/item?id=36683756 https://news.ycombinator.com/item?id=36683756

metadat 3y ago
Expanded version of previous recent coverage and discussions: Intel Exiting the PC Business as It Stops Investment in the Intel NUC https://news.ycombinator.com/item?id=36680674 https://news.ycombinator.com/item?id=36680674 (127 comments) Intel is quitting on NUC computers https://news.ycombinator.com/item?id=36683756 https://news.ycombinator.com/item?id=36683756 (3 comments)
dang 3y ago
Comments moved to https://news.ycombinator.com/item?id=36680674 https://news.ycombinator.com/item?id=36680674. Thanks!
- nubinetwork 3y ago
  While I have your attention, is there any chance we can get a dupe checker before posting? Cheers Edit: case in point, the Sarah Silverman suing openai posts, the Evernote posts, the fairphone posts, the France phone spying posts, the Canadian thumbs up emoji posts, I could go on sadly...
  - dang 3y ago
    HN has a dupe checker based on URLs. If you're asking for a dupe checker based on content, not URLs, that's harder than it sounds. I don't know how to do it and I don't know of anyone who's solved the problem. To solve it, you don't just have to identify when two articles contain the same story—you also have to distinguish when they don't contain the same story, even if the content is similar. Since we don't know how to do that, we rely on a combination of moderation and community feedback (i.e., human judgment) to identify when a story has already had significant attention on HN, and merge or downweight the follow-ups. Since it isn't automated, this approach misses things. It does catch most of them though.
    - nubinetwork 3y ago
      This is hn, just throw some ai at it...
      - dang 3y ago
        The relevant data is all public, if you or anyone want to take a crack at it.

https://news.ycombinator.com/item?id=36680674 https://news.ycombinator.com/item?id=36680674 https://news.ycombinator.com/item?id=36683756 https://news.ycombinator.com/item?id=36683756

metadat 3y ago

Expanded version of previous recent coverage and discussions: Intel Exiting the PC Business as It Stops Investment in the Intel NUC https://news.ycombinator.com/item?id=36680674 https://news.ycombinator.com/item?id=36680674 (127 comments) Intel is quitting on NUC computers https://news.ycombinator.com/item?id=36683756 https://news.ycombinator.com/item?id=36683756 (3 comments)

dang 3y ago

Comments moved to https://news.ycombinator.com/item?id=36680674 https://news.ycombinator.com/item?id=36680674. Thanks!

nubinetwork 3y ago
While I have your attention, is there any chance we can get a dupe checker before posting? Cheers Edit: case in point, the Sarah Silverman suing openai posts, the Evernote posts, the fairphone posts, the France phone spying posts, the Canadian thumbs up emoji posts, I could go on sadly...
- dang 3y ago
  HN has a dupe checker based on URLs. If you're asking for a dupe checker based on content, not URLs, that's harder than it sounds. I don't know how to do it and I don't know of anyone who's solved the problem. To solve it, you don't just have to identify when two articles contain the same story—you also have to distinguish when they don't contain the same story, even if the content is similar. Since we don't know how to do that, we rely on a combination of moderation and community feedback (i.e., human judgment) to identify when a story has already had significant attention on HN, and merge or downweight the follow-ups. Since it isn't automated, this approach misses things. It does catch most of them though.
  - nubinetwork 3y ago
    This is hn, just throw some ai at it...
    - dang 3y ago
      The relevant data is all public, if you or anyone want to take a crack at it.