Kore: Binary File Format Optimized for Modern Data Systems (Open Source)

A shiny new open-source project called Kore just strutted onto the big-data stage promising eye-popping speed, smaller files, and easy use with Spark, a popular tool for processing huge piles of information. The pitch is pure swagger: way faster than old-school text formats, tighter compression, and battle-tested before release. One booster even arrived like a hype man at a product launch, bragging about three years of production testing, support across a bunch of programming languages, and even a VS Code viewer for .kore files. In other words: this isn’t just a file, it’s a whole vibe.

But the comments? Oh, they immediately turned into a trust issues support group. The biggest reaction was basically: cool story, now show the benchmarks. Several readers zeroed in on the awkward compression claim, asking why Kore boasts a lower percentage than Parquet when, to normal humans, a bigger number sounds better. Others wanted to know the real tradeoff: what do you gain, what do you lose, and can this thing handle truly giant datasets without falling over? One commenter called the whole thing a little “vibe-codey,” which is internet-speak for “this looks slick, but would I trust it with my paycheck?” Another piled on by asking for comparisons to rivals like Vortex and wishing for a DuckDB extension.

So yes, Kore launched with bold promises — but the real headline from the crowd is prove it, compare it, and explain it like we’re not already sold.