Arun Isaac <[email protected]> writes: >>> Codeberg doesn’t have (yet) the same deduplication that Github has, with >>> our 389 forks and ~500M per repository, that’s almost 200G! >> >> Which is... almost nothing? You can get 26 TB HDD for ~750 EUR. >> Average ThinkPad these days comes with 1 TB drive. > > It's less about the sheer size of the data, and more about the CPU and > I/O scaling complexity that comes with it. When GitHub is contacting > individual projects to manage the strain on their infrastructure, it's > serious. > > Here are two cases (Nix and Homebrew) where GitHub has complained about > the strain on their servers. > > https://discourse.nixos.org/t/nixpkgs-core-team-update-2025-11-30-github-scaling-issues/72709 > https://github.com/Homebrew/brew/pull/9383
Well, since Codeberg does not do any deduplication (yet), each repository can be considered in isolation, and, realistically speaking, most of the forks will be quiet with low traffic. I would expect that to simplify the situation a bit. > > And, here's a hypothetical article by the original author considering > the merits of implementing Git on top of Postgres. > > https://nesbitt.io/2026/02/26/git-in-postgres.html > This looks neat. I somewhat like the current git design (I can just rsync files around, and do not have to figure out how to run postgres first), but the article is an interesting read nevertheless. Tomas -- There are only two hard things in Computer Science: cache invalidation, naming things and off-by-one errors.
signature.asc
Description: PGP signature
