From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <e@80x24.org>
X-Spam-Checker-Version: SpamAssassin 3.3.2 (2011-06-06) on dcvr.yhbt.net
X-Spam-Level: 
X-Spam-ASN: 
X-Spam-Status: No, score=-3.2 required=3.0 tests=ALL_TRUSTED,AWL,BAYES_00
 shortcircuit=no autolearn=unavailable version=3.3.2
X-Original-To: yahns-public@yhbt.net
Received: from localhost (dcvr.yhbt.net [127.0.0.1]) by dcvr.yhbt.net
 (Postfix) with ESMTP id 3A5EC1F980; Fri, 29 Apr 2016 07:35:08 +0000 (UTC)
Date: Fri, 29 Apr 2016 07:35:08 +0000
From: Eric Wong <e@80x24.org>
To: yahns-public@yhbt.net
Subject: corking headers on Transfer-Encoding:chunked
Message-ID: <20160429073508.GA30698@dcvr.yhbt.net>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
List-Id: <yahns-public@yhbt.net>

I don't think I can safely apply the following patch,
at least I don't think we can do it by default.
I wish I could, given how I always use Rack::Deflater:

--- a/lib/yahns/http_response.rb
+++ b/lib/yahns/http_response.rb
@@ -152,6 +152,9 @@ def http_response_write(status, headers, body)
         when %r{\AContent-Length\z}i
           flags |= MSG_MORE if have_more?(value)
           kv_str(buf, key, value)
+        when %r{\ATransfer-Encoding\z}i && value =~ /\bchunked\b/i
+          flags |= MSG_MORE
+          kv_str(buf, key, value)
         when "rack.hijack"
           hijack = value
         else

Problem is, we never know how long it'll take an app to
generate the first part of the response body in:

	body.each { |chunk| ... }

And it is likely somebody might want to send a header out
ASAP, making the 200ms MSG_MORE delay unnacceptable; right?