URL-backed binary storage
Why large app files, media and OS releases use the binary URL store instead of the ordinary JSON payload path.
currentstatus
Storage developers · Platform developersaudience
School-GIT docssource
SCHX nativeruntime
SCHX / FEATURE LAYER
Source map
- lib/cdn/url-db/url-binary-store.ts
- lib/cdn/url-db/global-cookie-system.ts
SCHX / FEATURE LAYER
Binary records
The binary store persists large byte streams as authenticated URL-backed chunks and stores compact metadata containing identity, length, MIME type and SHA-256 information.
Higher-level app and OS records keep references to those binary objects rather than embedding large base64 files in the ordinary metadata record.
SCHX / FEATURE LAYER
Immutable release use
Web Bro app versions and School-GIT OS releases point at immutable file descriptors. Download routes stream those exact records to the client, where they are hash-verified again.
SCHX / FEATURE LAYER
References are not authorization
A record identifier is a locator inside the platform’s storage model. Authentication/ownership checks still belong at the API layer.
SCHX / FEATURE LAYER