The Forge We Deserve

(btao.org)

23 points | by icy 2 hours ago

10 comments

  • OhSoHumble 1 hour ago
    I personally just self-host. The project that I'm working on right now uses Gitea internally with an action runner.

    I use Digital Ocean and run a Nomad cluster on top of that. Gitea and its action runner builds container images and then pushes nomad job definitions to the cluster. I have zero downtime deployment and rolling deploys. Dagger.io is in there somewhere to make local CI mirror what happens in the action runner.

    Setting up Gitea was honestly just an afternoon of work. The sqlite database is backed up to Cloudflare R2 and the code is mirrored. It's unlikely that my project will take off to the point that I'll need to upgrade Gitea to something else, but it's extremely easy to setup, maintain, and build CI/CD on top of.

    • notpushkin 38 minutes ago
      Codeberg runs on Forgejo (a Gitea fork) and is fairly big, so you’re in a good company. Maybe you’ll need to swap in another database, but SQLite is surprisingly powerful, too.

      While there are some problems with open source projects self-hosting their forges without federation, that’s really not the end of the world IMO (but lots of love for Tangled nevertheless).

      • KronisLV 15 minutes ago
        Also running Gitea, pretty pleasant experience!

        Moved over from Sonatype Nexus to Gitea Packages as well since administering Nexus is annoying: https://docs.gitea.com/usage/packages

        However for CI I use WoodpeckerCI (previously used Drone but migrated over), it works well with containers and is delightfully simple: https://woodpecker-ci.org/

        Though I guess in my case I don't collaborate with others much outside of work, so it's just something to interact with across computers and servers.

  • znnajdla 15 minutes ago
    Very timely as I just set up Forgejo yesterday. There were immediate unexpected benefits moving off GitHub. It’s hosted on an Orbstack VM on a Mac Mini M4 on my local Tailscale network and the UI is incredibly snappy and quick to browse. Because it’s located inside my tailnet with no public access I don’t need to setup authentication, so agents running inside our network can clone a repo with no fuss (I just protect master branch pushes with a config rule). Github auth is a pain to securely distribute, when you self-host the problem just disappears. Also because it’s running inside our tailnet we can use Forgejo actions for infrastructure-related tasks: Ansible scripts that provision or maintain Mac Minis or Linux servers on the same tailnet over Tailscale SSH. Much simpler than Kubernetes which would be overkill for a small homelab/startup. And the Orbstack VM is literally two clicks to backup and restore or move to another host — I setup the VM running Forgejo on my local MacBook then a few minutes later exported the VM image and transferred it to a Mac Mini server in a few clicks. Backup and restore an entire server to a single file — that is the fastest and easiest server migration I ever performed. I literally can’t think of a single reason I would want to go back to GitHub. Forgejo is a joy to use.
  • preya2k 46 minutes ago
    It seems ironic, that in the social media space, AT protocol based instances are basically centralised (Bluesky) and ActivityPub based instances (Fediverse, Mastodon) have a much healthier grade of federation.

    Whereas in the "forges" space, it seems Tangled drives federation forward much faster than the ActivityPub-based federation features of Forgejo/Gitea (which are progressing really slow).

  • bestouff 1 hour ago
    I'm way more hopeful in a fully decentralized protocol like Forgefed* than some AT still-centralized thing.

    * https://forgefed.org/

    • icy 1 hour ago
      Cool. It's been years since ForgeFed came out but it isn't anywhere nearly as federated as Tangled is. What gives?
      • notpushkin 52 minutes ago
        It’s funny that Bluesky is nowhere nearly as federated as Mastodon btw. (I guess people looking for Twitter replacement don’t really care to run their own PDS (or even know they have the option), while developers like tinkering with that kind of stuff.)

        No idea what went wrong with ForgeFed, yeah.

        ---

        Edit:

        > Cool. It's been years since ForgeFed came out but it isn't anywhere nearly as federated as Tangled is. What gives?

        Just read your other comments. Sorry, but I must point out that it would be polite to disclose that you’re working on Tangled, especially when posting takes like this on a competitor technology.

    • Gualdrapo 43 minutes ago
      Just tried the link to Vervis, Firefox mobile yells about a SSL error, went around it and still it returned a 404...

      (but what I'd know, I thought this was about typography forges - didn't know there were also software forges...)

  • sshine 39 minutes ago
    Tangled and Radicle are both really cool, but add too much mental gymnastics compared to just running Forgejo.

    What I like about (the idea of) ForgeFed is that it lets existing forges speak to each other.

    In practice I probably just need Forgejo and GitLab to be able to speak to each other.

    I believe the future of GitHub, for me, is to solve two problems:

      - Discoverability for public open-source projects
      - Backup since self-hosting is fragile long-term
    
    So many times when I try to visit the source code of some package uploaded to crates.io, the self-hosted git no longer exists.

    GitHub repos sit stale for decades.

    For day-to-day reliance, my self-hosted Forgejo and CI runners have better uptime.

    Only pet peeve with Forgejo:

      - It's a highly active project, RFCs, tons of PRs and issues.
      - Becoming a daily user, I want to extend it, and in its beautiful simplicity, it's not highly extensible.
      - So to avoid maintaining a fork of a very active project, extending it in unison is a social commitment.
    
    What a luxury problem, but still.

    I'd like to see more hosted Forgejo solutions pop up; it's very low-resource cost.

    • IshKebab 3 minutes ago
      I agree, these things just seem to add too much mental complexity compared to the advantages. I think the main attractive thing about Tangled is that it supports proper stacked PRs. But on the other hand it doesn't support private repos at all, and Github is getting stacked PR support soon (fucking finally)...

      It's hard to see the advantage of Tangled over Codeberg for example.

  • throwwwll 1 hour ago
    The forge we deserve is the one that is built for jj and local review.
  • MadsRC 1 hour ago
    I would love to see a forge that consists of composable components, self-host able with federated identities and maybe a few centralized indexers
    • icy 1 hour ago
      That's Tangled! Self-host your knots, AT Protocol for identities, and an "indexer" at tangled.org (easily replaced).
  • hkt 1 hour ago
    I'm pretty sure the actual forge we deserve at this point is one that is a membership organisation, eg, owned by its (paying) members.

    Members elect the board which chooses the CEO. A cooperative, in other words. The tech is a solved problem, with lots of open source around to do it. Enough members means paid operations and development staff, or outsourcing one or both, or grants to open source devs, etc. The possibility is there.

    That's how we prevent cultural drift: by actually controlling the company.

    • notpushkin 32 minutes ago
      > I'm pretty sure the actual forge we deserve at this point is one that is a membership organisation, eg, owned by its (paying) members.

      Like https://codeberg.org/? :-)

    • zero-st4rs 9 minutes ago
      hear hear!
  • BrenBarn 1 hour ago
    > So, what’s next? There are a lot of Git forges out there.

    How about some non-Git forges (like Mercurial forges)? That's what I'd like to see.

    • patates 1 hour ago
      I miss mercurial sometimes, but we have so much tooling around git these days...
  • lloydatkinson 1 hour ago
    > Or perhaps everyone moves to some hot-new-AI-first-forge, and then we go through the same cycle of enshittification in 10-20 years.

    I don't think it would take years for an AI first platform to enshittify, it would be instant.