A save() that takes the state lock over logs nothing, so a stale window overwriting a newer state.json cannot be traced #1041

Owner: Dev3

Found while landing #1027 / PR #1032 (review comment: #1032 (comment)).

Reproduce

refresh() prints a line when it takes the lock over from a host that has closed (core/src/host/store.js:400-408, "the other dsh has closed; this one now saves the Bot team"). The save() path added in #1027 takes the lock over through the same claimLock() (:434-441) but prints nothing at all, and warnReadOnly() (:410) is only reached when the takeover fails.

Why it matters

The case #1027 fixes is exactly "a window that showed the Bot team read-only writes over a state.json another window has updated meanwhile". When that happens the write is silent: there is no line saying a takeover occurred, which pid/host held it before, or that the file on disk had changed since this window's last read. A user reporting lost data has nothing to grep for, and we cannot tell a legitimate takeover from a stale window that overwrote a newer file.

Acceptance

• a save() that takes the lock over from a holder that has closed emits exactly one line naming the previous holder's pid and host (same shape as the refresh() line, or a distinct wording — the repository is English);
• the line is asserted by a test in core/test/home-lock.test.js (capture the logger/warn sink the same way the existing #14 read-only test does);
• the successful path stays quiet when this window already held the lock, and a live holder still returns it read-only (Read-only host: flybot/* reads rejected, social timer ignores writer lock (PR #2 review) #14 unchanged).

Scope is the log line, not a change of behaviour.