- From: Martin Thomson <martin.thomson@gmail.com>
- Date: Thu, 10 Apr 2014 15:25:57 -0700
- To: Daniel Sommermann <dcsommer@fb.com>
- Cc: HTTP Working Group <ietf-http-wg@w3.org>
On 10 April 2014 15:17, Daniel Sommermann <dcsommer@fb.com> wrote: > Should the language read that the client SHOULD retry, subject to some > rate-limiting? A client is never obligated to do anything. Use of "SHOULD" would imply an obligation of the sort where there are negative consequences for some party if they decided not to. In this case, the only entity affected by a client's refusal to retry is the client, so I don't see any interoperability impact to the choice. http://http2.github.io/http2-spec/#Reliability does make it clear that if the client still intends to make the request, it can.
Received on Thursday, 10 April 2014 22:26:27 UTC