Not mine, just cross-posting.

  • dan@upvote.au
    link
    fedilink
    arrow-up
    0
    ·
    1 year ago

    Yeah, I’d guess the same. I wonder how well that’d scale as the number of instances grows larger and larger…

    • xavier666@lemm.ee
      link
      fedilink
      arrow-up
      1
      ·
      1 year ago

      If my graph-theory basics are still correct, I think it will scale linearly with more instances. The process can take longer to compute, but it’s not like the process will take more compute resources.

      However, the more interesting problem would be if a random instance is created which is only connected to other newer instance, do we need to start the calculation from scratch or is there a “mid-point” from we can start the search?