Write RSOS three ways.
Compose a program as a painting, key it in with the RSOS keyboard, or write it as text and compile it. All three produce the same language and run through the same compiler checks.
The painting editor
Place objects in a scene, connect them, colour them and set their order. The editor compiles the painting to RSOS and back, so you can move between the picture and the text.
Open the editorGood first steps
- Load an example and look at how each object becomes a statement.
- Move one object to another room and watch the compiled program change.
- Remove a consent line from a person-facing program to see the compiler refuse it (E018).
- Pick a speed-test task and watch the REACH panel: the task as a stretch target, the measured distance after each gesture, and a prompt to name what stops you.
The RSOS keyboard
One key or chord emits a whole component, operator, perspective, state, depth, order or direction, so a program goes in at the speed it is seen and no required slot can be left out silently. Practice mode runs short sessions and draws your learning curve.
Open the keyboardHow it was checked
Twelve targets were keyed by a headless browser, compiled and decoded blind from the code alone. The first run passed 8 of 12 and found one bug (an operator key overwrote a finished model). After the fix: 12 of 12, with 88 of 89 fields recovered (0.989). The human test, HT2, follows a learning phase.
Text and the compiler
Install from the repository and check any example. The compiler reports errors and warnings by code, including the consent and harm checks.
An MCP server (pip install rsoslang, then rsoslang mcp) lets an AI assistant compile and run RSOS
programs as tools.
# get the code
git clone https://github.com/r-a-elu-architecture/RSOS-Language
cd RSOS-Language
pip install -e .
# check and compile an example
python -m rsoslang check examples/robot_head.rsl
python -m rsoslang ir examples/sol_remote.rsl
# reproduce every result
./reproduce.sh