W3C home > Mailing lists > Public > ietf-http-wg@w3.org > January to March 2011

Re: Content-Disposition: *sender* advice

From: Julian Reschke <julian.reschke@gmx.de>
Date: Thu, 24 Feb 2011 08:54:41 +0100
Message-ID: <4D660EC1.3040502@gmx.de>
To: Mark Nottingham <mnot@mnot.net>
CC: HTTP Working Group <ietf-http-wg@w3.org>
On 24.02.2011 06:39, Mark Nottingham wrote:
> I wouldn't say we break FireFox by recommending this ("breaking" means that C-D doesn't work at all, or produces garbage), and I wouldn't say they've been doing the right thing (because they implemented to prefer the non-internationalised filename).

When I said "break" I mean that sending a I18Nized filename will not 
have the desired effect.

When I said "do the right thing" I mean that *they* implement the 
published specs properly (at least with respect to this), while other 
UAs do not (Safari), or just started to do so (Chrome 9, IE9).

The recommendation to pick "filename*" over "filename" is new (as of 
this spec), and it's only a SHOULD (we could change that, though).

> Also, based on their bug traffic, they unfortunately don't seem too eager to fix this.

Unfortunately. Hopefully the plan with respect to making major releases 
more frequent will happen; in which case having this in FF5 wouldn't be 
too bad.

> When you say "no advice" what do you mean?

I mean not having this appendix at all (after all, it appears it never 
was discussed seriously until last week, and we are months past WG LC), 
or alternatively to have a short statement that producing I18Nized field 
values is tricky because of the various implementation problems, and 
then have a pointer to something that's easier to maintain.

Maybe we could go to IETF LC and mark this question as something we're 
looking for feedback for?

Best regards, Julian
Received on Thursday, 24 February 2011 07:55:23 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Friday, 27 April 2012 06:51:37 GMT