Skip to main content

Compress — the HZC compressor

Compress is the framework's file compressor. It does the ordinary job — make a file smaller — and two things an ordinary compressor doesn't: it stamps every archive to the clock, and it measures what it's compressing. Compression here is also a diagnostic.

Lossless and lossy, in one place​

You can compress losslessly — reversible, byte-for-byte — or lossily where the data tolerates it, trading fidelity for size. Both produce the framework's own archive format, .zeqc (Zstd under the hood), and both report the exact compression ratio they achieved (compressed ÷ original), so you're never guessing what you saved. The archive carries the Zeqond it was made at, so an archive is dated on the framework's clock rather than a wall clock.

Compression as measurement​

Because a compressor's whole job is to remove redundancy, how well something compresses tells you what it is. Compress surfaces that directly: it computes the Shannon entropy of your bytes and classifies them:

EntropyClassificationMeaning
< 3.5RAWstructured, redundant data — compresses far
3.5 – 6COMPRESSEDalready dense; little left to squeeze
> 6ENCRYPTEDnear-random; no redundancy to remove

That's a genuinely useful signal: point Compress at a blob and it tells you whether you're looking at raw data, something already packed, or HITE-style ciphertext — before you've decoded a thing.

It's a building block​

Compress runs client-side in the browser, but it's the same primitive you can drive from your own app or a contract: compress an artifact before you seal it, or read the entropy classification into a pipeline that routes raw and encrypted data differently. Like every app here, it's a face on the framework, not a silo.

Open it​

Open Compress →

Compress and decompress files in the browser, with the ratio and entropy read shown live.