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

Section 6.9 (WINDOW_UPDATE), shouldn't it always be a connection error if window increment is 0?

From: Matthew Cox <macox@microsoft.com>
Date: Tue, 5 Aug 2014 16:33:36 +0000
To: "ietf-http-wg@w3.org" <ietf-http-wg@w3.org>
Message-ID: <0d22d7944f2a41d0979070a5b202534e@BY2PR03MB521.namprd03.prod.outlook.com>
Draft 14 adds this paragraph to section 6.9:

   A receiver MUST treat the recipt of a WINDOW_UPDATE frame with an

   flow control window increment of 0 as a stream error (Section 5.4.2<http://tools.ietf.org/html/draft-ietf-httpbis-http2-14#section-5.4.2>)

   of type PROTOCOL_ERROR; errors on the connection flow control window
   MUST be treated as a connection error (Section 5.4.1<http://tools.ietf.org/html/draft-ietf-httpbis-http2-14#section-5.4.1>).

The only allowed values are 1 to 2^31 - 1.  This would lead me to believe that this should always be a connection error instead of being based on the stream id in the frame.

I don't see any reason to allow a server to send a bogus frame in this way even if they set a stream id.

Unless someone has a good reason to allow this I purpose making this always a connection error regardless of the intended target of the window update.


Received on Tuesday, 5 August 2014 16:34:20 UTC

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