Show HN: Albedo – single-file listenable document database in Zig https://ift.tt/Y3ulKdQ

Show HN: Albedo – single-file listenable document database in Zig It's literally an indexable BSON array with query listening, replication and basic crash recovery. While working on my little projects I always felt a need for some sort of simple storage, where I can just have a single file. However sqlite with its' schemas and migrations always felt icky, like an overkill to have a single file db, but a whole folder set for migrations. JSON files would require to be loaded all at once and for something small but thick, like microservices that would hold their own state this would be a startup time cost. So idea came initially as an "indexable and partially loadable json object array". From the very beginning I've wrote it in zig, unlike my other projects where I flip flop with implementation language, for the sole reason that I wanted to be able to use it in as many places as physically possible. So it was something on level of C\C++, Rust and Zig. C langs are awesome, but I genuinely didn't enjoy them. After initial tries Rust felt overbearing, so by method of exclusion I landed on the lizard lang. I've been working on it for about a year, all while learning zig, relearning working with memory and adding new features. This whole time I've been mainly comparing it to and learning so much from sqlite, and eventually, after tuning, I feel like the performance is close enough for me (at least in my very basic benchmarks on my laptop). Now it's my personal little engine I use in my mobile apps through flutter plugin, in my small backends using bun ffi \ node native modules. Would be cool if it would solve someone's else pains. Happy to answer any questions https://ift.tt/rWhBO6A August 10, 2026 at 12:03AM