fesk-rt

FESK decoding, real time ⚡

Usage

This decoder is meant to decode messages created with the FESK TX library that ships with the Second Movement firmware for Sensor Watch.

If you want to test the out of the box demo mode, enable the watch face fesk_demo_lite_face in your watch build and flash the resulting UF2 to your Sensor Watch as described in the official hardware documentation. Once the face is running you can press the alarm button on the watch to broadcast a test FESK frame, and this can be decoded in the web app.

Or you would use the library in your own watch face and broadcast whatever you want. Something like:

typedef struct {
   fesk_session_t session;
} fesk_demo_lite_state_t;

       case EVENT_ALARM_BUTTON_UP:
           {
               fesk_session_config_t config = fesk_session_config_defaults();
               config.static_message = "test";
               fesk_session_init(&state->session, &config);
               fesk_session_start(&state->session);
           }
           break;

At the time of writing the FESK TX library is not merged, but can be found here: https://github.com/joeycastillo/second-movement/pull/139

Supported character codes

The decoder recognises the following 6-bit codes. The table lists every possible pattern along with the symbol it represents or how it is reserved.

| Bits | Decimal | Symbol | Notes | | ------- | ------- | ----------------- | ----------------------------- | | 000000 | 0 | a | | | 000001 | 1 | b | | | 000010 | 2 | c | | | 000011 | 3 | d | | | 000100 | 4 | e | | | 000101 | 5 | f | | | 000110 | 6 | g | | | 000111 | 7 | h | | | 001000 | 8 | i | | | 001001 | 9 | j | | | 001010 | 10 | k | | | 001011 | 11 | l | | | 001100 | 12 | m | | | 001101 | 13 | n | | | 001110 | 14 | o | | | 001111 | 15 | p | | | 010000 | 16 | q | | | 010001 | 17 | r | | | 010010 | 18 | s | | | 010011 | 19 | t | | | 010100 | 20 | u | | | 010101 | 21 | v | | | 010110 | 22 | w | | | 010111 | 23 | x | | | 011000 | 24 | y | | | 011001 | 25 | z | | | 011010 | 26 | 0 | | | 011011 | 27 | 1 | | | 011100 | 28 | 2 | | | 011101 | 29 | 3 | | | 011110 | 30 | 4 | | | 011111 | 31 | 5 | | | 100000 | 32 | 6 | | | 100001 | 33 | 7 | | | 100010 | 34 | 8 | | | 100011 | 35 | 9 | | | 100100 | 36 | Space ( ) | | | 100101 | 37 | , | | | 100110 | 38 | : | | | 100111 | 39 | ' | | | 101000 | 40 | " | | | 101001 | 41 | Line feed (\n) | | | 101010 | 42 | — | Reserved for future expansion | | 101011 | 43 | — | Reserved for future expansion | | 101100 | 44 | — | Reserved for future expansion | | 101101 | 45 | — | Reserved for future expansion | | 101110 | 46 | — | Reserved for future expansion | | 101111 | 47 | — | Reserved for future expansion | | 110000 | 48 | — | Reserved for future expansion | | 110001 | 49 | — | Reserved for future expansion | | 110010 | 50 | — | Reserved for future expansion | | 110011 | 51 | — | Reserved for future expansion | | 110100 | 52 | — | Reserved for future expansion | | 110101 | 53 | — | Reserved for future expansion | | 110110 | 54 | — | Reserved for future expansion | | 110111 | 55 | — | Reserved for future expansion | | 111000 | 56 | — | Reserved for future expansion | | 111001 | 57 | — | Reserved for future expansion | | 111010 | 58 | — | Reserved for future expansion | | 111011 | 59 | — | Reserved for future expansion | | 111100 | 60 | — | Reserved for future expansion | | 111101 | 61 | — | Reserved for future expansion | | 111110 | 62 | — | Frame start marker | | 111111 | 63 | — | Frame end marker |

Codes 42 through 61 are reserved for future expansion and should not appear in normal frames. Codes 62 and 63 mark frame boundaries and are not emitted as characters.

Why "FESK"?

FESK is a word play on FSK (Frequency-shift keying). The FESK protocol itself supports binary FSK (BFSK) and 4FSK. But the word “fesk” means “fish” in northern Norwegian dialect. Coincidentally, in Norway, I would say one can also use the word as a mild swear word. Something like “what the fish” or “oh… fish!”

Built with AI Assistance

This project is crafted with help from AI tools but steered with human supervision. OpenAI's gpt-5-codex model assists with code, code reviews and implementation planning, and Anthropic's Claude 4.5 Sonnet model has provided complementary insights, debugging and code during those same review, planning and coding sessions. It's very, very far from a one shot prompt 😅

Licence

MIT