• gencha@lemm.ee
    link
    fedilink
    arrow-up
    59
    arrow-down
    6
    ·
    4 months ago

    I feel like most people base their decision on license purely on anecdotes of a handful of cases where the outcome was not how they would have wanted it. Yet, most people will never be in that spot, because they don’t have anything that anyone would want to consume.

    If I had produced something of value I want to protect, I wouldn’t make it open in the first place. Every piece of your code will be used to feed LLMs, regardless of your license.

    It is perfectly fine to slap MIT on your JavaScript widget and let some junior in some shop use it to get their project done. Makes people’s life easier, and you don’t want to sue anyone anyway in case of license violations.

    If you’re building a kernel module for a TCP reimplementation which dramatically outperforms the current implementation, yeah, probably a different story

    • SuperSpruce@lemmy.zip
      link
      fedilink
      arrow-up
      20
      ·
      edit-2
      4 months ago

      I once read that the license should be smaller than your code. Gives me a good baseline:

      • Permissive license for small projects and little tests

      • Copyleft license for big projects

    • BaumGeist@lemmy.ml
      link
      fedilink
      arrow-up
      19
      ·
      4 months ago

      Well, ideally you’re choosing your license based on the cases where it differs from others and not the majority of times where it doesn’t make a difference.

      Someone aiming to make Free software should use a copyleft license that protects the four freedoms, instead of hoping people abide by the honor system.

      Also, no one can 100% accurately predict which of their projects will get big. Sure, a radical overhaul of TCP has good odds, but remember left-pad? Who could have foreseen that? Or maybe the TCP revision still never makes it big: QUIC and HTTP/3 are great ideas, and yet they are still struggling to unseat HTTP/2 as the worldwide standard.

      • lolcatnip@reddthat.com
        link
        fedilink
        English
        arrow-up
        8
        arrow-down
        1
        ·
        4 months ago

        Seems like using a copyleft on the reference implementation of a new protocol is a great way to ensure the protocol is never widely adopted.

      • gencha@lemm.ee
        link
        fedilink
        arrow-up
        1
        ·
        4 months ago

        People who used left-pad deserved everything that happened to them. But, very valid point.

        There is no honor system. If your code is open for commercial reuse, that’s it. If you have any expectations that are not in line with that, then yes pick a different license.

        I guess I agree with you, I’m just phrasing it from a different perspective.