I’ve read this comment on lemmy’s github repository by sbordeyne:

I closed it cause I have no interest in having to battle dessalines on minor nitpicks. Feel free to take the branch and reopen the pull request.

My thought process was that since lemmy is open source, we should be free to add new features without having to battle over philosophies, which was made apparent to me with his reviews and very strong opinions over which features to greenlight and which features to debate at length over (see infinite scrolling, and this very issue which was only approved on the condition we make it a setting that is off by default). I just don’t have the energy to contribute and be challenged on every little thing, I’ll just remain a plain user of the platform and not bother contributing to the project anymore.

Originally posted by sbordeyne in #282 (comment)

Instead of contributing to a project where only @dessalines@lemmy.ml, lionirdeadman and @nutomic@lemmy.ml dictate what gets approved, why not make a fork with the settings most people agree on?

My stance is that I wouldn’t contribute to a project hosted on GitHub, or a project that considers a few opinions more important than the majority.

Controversial takes
dessalines doesn't want SEO so we get URLs for robots

It might makes sense for the semantic web, but IMO it really doesn’t add that much value… look at how silly this type of thinking has gotten w/ reddit: https://www.reddit.com/r/aww/comments/glenz4/cute_baby_bunnies_think_the_golden_retriever_is/

Its gotten to reddit/community/comments/post_id/weird_lengthened_post_name, when it should be just reddit/post/postid

_Originally posted by @dessalines@lemmy.ml in https://github.com/LemmyNet/lemmy/issues/875#issuecomment-652165036_

:::

In my opinion, open-source projects should follow a set of principles to ensure transparency, community engagement, and effective development. Here are some key points that encapsulate these principles:

  • Clear Decision-Making Process for Future Releases: Open-source projects should have a well-defined process for deciding what features and improvements make it into each release. This process should be documented and accessible to the community. (For example, you can read about how Discourse decides what goes into each release here).

  • Complaint-Driven Development: Embrace the concept of “complaint-driven development,” where user feedback and complaints guide the prioritization of issues and features to address.

  • Transparent Donation Management: Be open about how donation money is allocated. Implementing a bounty system can further enhance transparency, enabling contributors to work on specific tasks for financial rewards.

  • Inclusive Community Involvement: Encourage active participation from the community in brainstorming and voting on potential features before finalizing the project roadmap. Community input can be invaluable in shaping the project’s direction.

  • Public Roadmap: Maintain a publicly accessible roadmap that outlines what is planned for upcoming releases. This roadmap should include the status of each feature or enhancement to keep users informed.

  • User-Friendly Feature Request Process: Provide a public feature request forum where users can submit their ideas. Clearly outline the process for how these requests will be reviewed and prioritized, ensuring transparency in the decision-making.

  • Effective Communication Channels: Avoid using disorganized or private communication platforms for project-related discussions. Instead, opt for well-structured, publicly accessible channels that don’t require users to log in to access information.

  • Document with a Blog: Utilize a blog or similar documentation platform to keep users and contributors informed about important project updates, changes, and developments.

  • star_nova@beehaw.org
    link
    fedilink
    English
    arrow-up
    37
    ·
    1 year ago

    A very direct answer to any “Why not just fork X?” question is, would you want to develop and maintain the fork in your free time and for no pay? Forking any project is deeper than just copying a repo and then making a few changes.

    This would involve finding a group of people intimately familiar with the codebase who also don’t mind committing all of their free time to development.

    When you fork a project, you also lose whatever payment/donation structure the original project had setup. Asking for some kind of payment to develop a larger project should not come as a surprise. You want to be able to attract halfway decent developers and they want their work to be valued. Beyond just approving PRs, you need a team to verify additions and work out all the conflicts and errors that arise.

    There very well may be a lot of settings people all agree on. But who is going to bite the bullet and put in all the time to add those features, potentially for the fork to be abandoned?

    I hope this doesn’t sound condescending because it’s not intended to be. I’m just trying to explain that the scope of “Just do this” is a lot larger than what it first seems.

    • jarfil@beehaw.org
      link
      fedilink
      arrow-up
      3
      ·
      1 year ago

      would you want to develop and maintain the fork in your free time and for no pay

      I hereby offer myself to maintain a fork for a minimal pay. Any takers?

      • wahming@monyet.cc
        link
        fedilink
        English
        arrow-up
        1
        ·
        1 year ago

        You can get it started as your first step, and see how many were serous about joining in