The diversion catches output that is actually flushed to stdout while
serving. A print() in a block-buffered process usually stays unflushed
in sys.stdout's user-space buffer, and the interpreter's exit flush
drains it onto the restored protocol stream, so the docs now say so
instead of implying every stray print is diverted. The logging advice
is unchanged: its handler flushes each record.
No-Verification-Needed: docs-only wording change
While serving on the process's real stdout, stdio_server now moves the
protocol pipe to a private descriptor and points fd 1 - and, on
Windows, the standard output handle - at stderr, restoring it when the
transport exits. A stray print() in handler code or a child process
writing to its inherited stdout lands in the client's log instead of
corrupting the JSON-RPC stream. The null device stands in when stderr
is unusable or, on POSIX, is detected as merged into stdout (2>&1).
The stdin claim generalizes into the shared _claim_fd mechanism: one
lock-guarded sentinel table covers both descriptors, private wire
duplicates are forced above the standard descriptor range so a process
started with a standard descriptor closed cannot hand the wire out as
its "stderr", and a failed claim degrades to serving the sys stream's
buffer in place exactly as v1 did.
Docs now describe the guarded behavior with its remaining gaps (output
flushed before serving begins, injected streams, merged stderr on
Windows), and the transport:stdio:stream-purity divergence narrows
accordingly.
Max Isbey