- From: Adam Barth <ietf@adambarth.com>
- Date: Wed, 15 Dec 2010 22:27:31 -0800
- To: Mark Nottingham <mnot@mnot.net>
- Cc: Julian Reschke <julian.reschke@gmx.de>, "Jungshik Shin (½ÅÁ¤½Ä, ãéïÙã×)" <jungshik@google.com>, Maciej Stachowiak <mjs@apple.com>, HTTP Working Group <ietf-http-wg@w3.org>
Why doesn't it affect the spec? It seems like implementations won't want to remove support for a feature that's used on a site as popular as Gmail. Adam On Wed, Dec 15, 2010 at 7:36 PM, Mark Nottingham <mnot@mnot.net> wrote: > Perhaps, but it depends on how Google sniffs the UAs. And I don't think it affects the spec. > > > On 16/12/2010, at 9:40 AM, Julian Reschke wrote: > >> On 15.12.2010 21:19, Jungshik Shin (½ÅÁ¤½Ä, ãéïÙã×) wrote: >>> ... >>> Before rushing to remove it (as an optional 'fallback') , I'd like to >>> have some 'numbers' about what web servers do (FYI, some Google products >>> emit RFC 2047 for Firefox and Chrome at the moment, but I guess Google >>> has to switch over to RFC 5987 for Firefox and Chrome). I'm not sure >>> whether the cost of supporting it is larger than the benefit. >>> ... >> >> Indeed. GMail seems to use RFC2047-encoding (when saving an attachment with non-ASCII characters in the filename) for Firefox (and likely for Chrome as well). >> >> So it's unlikely that UAs can remove this until this get fixed. >> >> Best regards, Julian >> > > -- > Mark Nottingham http://www.mnot.net/ > > > >
Received on Thursday, 16 December 2010 06:28:36 UTC