I think this will probably fail if it is positioned as a GitHub replacement.
As you said, Fossil may provide a more unified context for AI. But GitHub’s biggest advantage is that it effectively functions as a developer portfolio.
Federation is not really a buying reason for most developers. The core question is exposure. Git itself is distributed, so why did everything centralize around GitHub? The same thing happened with npm. Centralization is, in practice, something users often want.
So I think the real opportunity is not simply “GitHub should be replaced.” It is to target moments when there is already public pressure or momentum for teams to leave GitHub, and then focus on larger projects or teams that might actually move.
I suppose this is a kind of herd effect. Honestly, my main concern is whether the large teams capable of moving others will actually overcome GitHub’s inertia.
The idea is good, but I do not believe that the better technical solution necessarily wins. Also, from what I have observed, many open-source teams are actually hostile to the idea of AI context, so advertising the AI angle too strongly might even hurt adoption.
I think the killer feature matters more. If this is simply “hosted Fossil,” then the question becomes: what concrete problem does it solve?
At least from my perspective, your target audience and your proposed killer feature, AI-friendly context, do not seem fully aligned.
This is a nice improvement over GitHub, especially in the context of agents. Having code, issues, history, documentation, and discussion in a more self-contained and structured project format seems like it could make it much easier for agents to understand and work with the full context of a project.
I think this will probably fail if it is positioned as a GitHub replacement.
As you said, Fossil may provide a more unified context for AI. But GitHub’s biggest advantage is that it effectively functions as a developer portfolio.
Federation is not really a buying reason for most developers. The core question is exposure. Git itself is distributed, so why did everything centralize around GitHub? The same thing happened with npm. Centralization is, in practice, something users often want.
So I think the real opportunity is not simply “GitHub should be replaced.” It is to target moments when there is already public pressure or momentum for teams to leave GitHub, and then focus on larger projects or teams that might actually move.
I suppose this is a kind of herd effect. Honestly, my main concern is whether the large teams capable of moving others will actually overcome GitHub’s inertia.
The idea is good, but I do not believe that the better technical solution necessarily wins. Also, from what I have observed, many open-source teams are actually hostile to the idea of AI context, so advertising the AI angle too strongly might even hurt adoption.
I think the killer feature matters more. If this is simply “hosted Fossil,” then the question becomes: what concrete problem does it solve?
At least from my perspective, your target audience and your proposed killer feature, AI-friendly context, do not seem fully aligned.
Thanks for the thoughtful feedback, looks like I have some more thinking about how to frame this.
This is a nice improvement over GitHub, especially in the context of agents. Having code, issues, history, documentation, and discussion in a more self-contained and structured project format seems like it could make it much easier for agents to understand and work with the full context of a project.