The solution at the end was to detect IE via the user agent
IE's way for the filename and do the RFC 2231 way for others. But,
that's assumes UTF-8 URL encoding is turned on in IE.
Right, and that's what we did back then. That worked for FF users, IE
users in western countries, but not for Asian users.
It'd be nice if IE supported the RFC 2231 way.
And, to get there, I think it would be good if HTML5 stated that as a
The IE encoding is a lot better. In order to support clients using it in requests, I have to be able to parse the filename, and the IE syntax is much, much easier to parse than the 2231-based syntax. Why not file a bug report against IE so that it works all the time?
I also agree with the others that this isn't something that should be standardized in HTML, because it is not specific to HTML. I am implementing support for this (in both requests and responses) to my AtomPub implementations, for example. A seperate RFC for a *HTTP* Content-Disposition mechanism makes much more sense for use by non-HTML software. Make the IE syntax for the "filename" parameter the standard, and allow an additional "filename*" parameter for backwards-compatibility with UA's that implement the 2231 mechanism.