CydandClaude Fable 5
4fb91775fc
Plain-language explanation of the reply-latch wedge for non-technical pod owners, linked from the firmware README top. Frames the bug as a 'busy' sticky note the board forgets to take down on the give-up path, and how RIO 4.3 fixes it. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
1.9 KiB
Why your pod's stick sometimes "died" mid-battle — and why it won't anymore
A plain-language explanation for pod owners. The technical write-up is in RIOv4_2-ANALYSIS.md.
Inside every pod is a small computer board from 1994 (the RIO) whose job is to tell the game what you're doing: which buttons you pressed, and where you've moved the stick, throttle, and pedals.
The board and the game talk like two people on walkie-talkies. Every time the board reports your stick position, it waits for the game to answer "got it." During calm flying that works fine. But in a frantic firefight — lamps flashing, buttons hammering — the channel gets crowded, and once in a while that little "got it" gets drowned out.
When that happens, the board repeats itself a few times and then, sensibly, gives up and moves on. Here's the bug: while it's sending a position report, the board puts up a mental sticky note that says "busy — already sending one." When a report goes through normally, it takes the note down. But on the give-up path, the 1994 programmers forgot one line — the note never comes down. From that moment the board looks at the sticky note, decides it's still busy, and quietly throws away every stick and throttle update forever after.
That's the "wedge" you may have felt: mid-mission, your stick and throttle just stop doing anything, while the buttons and lamps still work fine — because those travel a different path that the sticky note doesn't block. Starting a new mission always cured it, since a new game tells the board to clean house, sticky note included.
The RIO 4.3 firmware adds the one missing instruction: whenever the board gives up on a report, it now takes the sticky note down too. We stress-tested it by mashing buttons at over nine presses per second for two minutes straight — the old board wedged under that abuse; the new one never misses a beat.