- From: Mark Nottingham <mnot@mnot.net>
- Date: Thu, 2 Sep 2010 11:14:17 +1000
- To: Henrik Nordström <henrik@henriknordstrom.net>
- Cc: Yves Lafon <ylafon@w3.org>, HTTP Working Group <ietf-http-wg@w3.org>
(just me) +1 On 30/08/2010, at 8:24 AM, Henrik Nordström wrote: > fre 2010-07-23 klockan 08:58 -0400 skrev Yves Lafon: > >> I don't see why it would be a MAY for a proxy and a SHOULD for a server, >> it seems better to keep them both as SHOULDs. > > Because a SHOULD means it's the recommended method, and as shown in this > case it's not a good general purpose recommendation. > > The optimization to have proxies reduce the traffic sent to the > requestor by eating the unwanted parts of the response when the server > responds with more data than requested should imho be a MAY, not a > SHOULD. > > For a server the situation is different as it has control over the > requested resource. > > Regards > Henrik > > -- Mark Nottingham http://www.mnot.net/
Received on Thursday, 2 September 2010 01:14:51 UTC