http: cork chunked responses for small savings

This only affects Linux users with MSG_MORE support.

We can avoid extra TCP overhead for sub-optimal chunk sizes
by using MSG_MORE even with chunk trailers under Linux.

This breaks real-time apps which require <= 200ms latency for
streaming small packets (e.g. implementing "tail -F"), but the
public-inbox WWW code does not (and will never) do such things.

diff --git a/lib/PublicInbox/HTTP.pm b/lib/PublicInbox/HTTP.pm
index c141fc896d0538875ca5a76c26c413ec910c4652..e19c592c99612ae29960676e5112166d5f5642eb 100644 (file)
--- a/lib/PublicInbox/HTTP.pm
+++ b/lib/PublicInbox/HTTP.pm
@@ -223,7 +223,10 @@ sub chunked_wcb ($) {
                return if $_[0] eq '';
                more($self, sprintf("%x\r\n", bytes::length($_[0])));
                more($self, $_[0]);
-               $self->write("\r\n");
+
+               # use $self->write("\n\n") if you care about real-time
+               # streaming responses, public-inbox WWW does not.
+               more($self, "\r\n");
        }
 }