W3C home > Mailing lists > Public > ietf-http-wg@w3.org > July to September 2014

(unknown charset) http2-14

From: (unknown charset) 山本和彦 <kazu@iij.ad.jp>
Date: Fri, 05 Sep 2014 09:27:15 +0900 (JST)
Message-Id: <20140905.092715.253318247027015671.kazu@iij.ad.jp>
To: (unknown charset) ietf-http-wg@w3.org
Hi,

I have read the http2-14 draft. Though I can accept historical reasons,
I would point out something.

1) SETTINGS_MAX_FRAME_SIZE should be SETTINGS_MAX_PAYLOAD_LENGTH.

   - Frame associates with both header and payload.
   - The name of frame header field is length, not size.

2) The origin of SETTINGS is 0x1 while other's origins are 0x0. I feel
   this is inconsistent.

3) Flags are frame-type local but I think the protocol designers try
   to make them global unique as much as possible:

	END_STREAM (0x1)
	ACK (0x1)
	END_HEADERS (0x4)
	PADDED (0x8)
	PRIORITY (0x20)

   I think that END_STREAM or ACK should be 0x2.

--Kazu
Received on Friday, 5 September 2014 00:27:42 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 30 March 2016 09:57:10 UTC