- From: Willy Tarreau <w@1wt.eu>
- Date: Thu, 7 Apr 2016 13:58:30 +0200
- To: Mark Nottingham <mnot@mnot.net>
- Cc: Matthew Kerwin <matthew@kerwin.net.au>, Piotr Jurkiewicz <pjurkiew@agh.edu.pl>, ietf-http-wg@w3.org
On Thu, Apr 07, 2016 at 08:06:22AM -0300, Mark Nottingham wrote: > The only thing I'll add is that there isn't any such constraint on registered > names in the RFC defining the registry. That doesn't mean that it's not > possible to argue for a name change, but I think there'd need to be a good > reason and strong supporting consensus, which seems to be lacking here. While I'm not fond of starting to use shortened tokens to describe slightly longer names, I'd remind that over the last 3 years we've been discussing a lot about how to shrink request sizes to save bytes on the wire and reduce page load time. This passes via reducing header field names and also token values. We need to start somewhere. Yes H2 did a great job in this area but that doesn't save us from being careful in all situatoins. From a technical point of view it seems to me that it makes sense to use a short name like "br". We won't run short of tokens for accept-encoding because of this. I think that a valid objection would be the discovery of some obscure implementation that already uses "br" and which would cause breakage. Regards, Willy
Received on Thursday, 7 April 2016 11:59:54 UTC