Experimental optical transport · v0.2

Data, woven into light.

Tjokode is a circular optical field built from continuous local waveforms rather than square barcode modules. This release makes generation fast, scans serially instead of overloading the camera client, and renders each recovered content type appropriately.

Fast vectorized rendererHybrid soft decodingType-aware results0.5 s temporal weaveText · links · files · tickets · contacts

Generate

The content editor now changes with the selected payload type.

Optical preview

A static result appears immediately. Slideshow mode returns PNG frames and a full-screen HTML player; MP4 is optional.

Choose a content type and generate a code.

Image recognition

Upload the original PNG, a photo, or an unordered set of slideshow captures.

Live camera client

One frame is processed at a time, preventing the overlapping decode requests that previously stalled scanning.

Fill the guide, keep all four anchors visible, and avoid screen glare.

Camera is stopped.

Static maximum

—

Implemented Ultra payload before application metadata.

Phone slideshow

—

Nominal optical payload at two images per second, before temporal repair.

V1 address space

—

A protocol ceiling, not a practical transfer recommendation.

Implemented profiles

Profile	Canvas	Petals	Pilots	Code rate	Static bytes	Bytes/s @ 0.5 s	Petal px

Can one code contain a megabyte?

Not while remaining an easy, single-shot phone scan. A one-megabyte payload is 8,388,608 information bits before locators, pilots, metadata, and error correction. Normal camera blur, demosaicing, perspective, and motion mean many sensor pixels are needed for each independent optical value.

Loading the camera-channel estimate…

Projected code/camera area	Discounted engineering bound	Interpretation

Architecture

The payload is not encoded as black-and-white modules. Each annular petal contains five signed spatial projections plus a two-bit Gray-mapped chroma angle. Distributed known petals calibrate the captured channel.

Tjokode symbol anatomy — Acquisition geometry and optical layers

PetalWave local channel — Seven-bit local analog microchannel

Decoder pipeline — Camera-to-payload recognition path

Temporal weave — 0.5-second erasure-recovery transport

Minimal Python API

from tjokode import generate, recognize
from tjokode.api import save_static

image = generate(
    "Hello from Tjokode",
    kind="text",
    profile="phone-robust",
    shape="hex",
)
save_static(image, "hello.png")

result = recognize("hello.png")
print(result.packet.text())