Re: [tlug] Firefox 3.0.1 doesn't respect <meta http-equiv="content-type">

[Curt Sampson <cjs@example.com>]

On 2008-09-12 11:13 +0900 (Fri), Stephen J. Turnbull wrote:

> What we claim is that the META element *may* be *useful* to the server
> in analyzing the contents of the file.

This point of view is not unreasonable. That said, I still think
that the mechanism is very poorly designed.

Essentially, it's mixing the problem of what the content of the file
might be with the issues of HTTP transport of the data in that file.
Additionally, it provides a way of specifying potentially conflicting
information, thus introducing the need to resolve those conflicts. We
could get rid of this by insisting that the server be responsible for
not sending conflicting information, but that would mean that any HTTP
server, to be compliant, needs to at the least be able to parse HTML.
Ouch.

A much more reasonable solution, to me, would be to entirely
disassociate any HTML tags from the vagaries of what might be used to
transport that HTML (HTTP or otherwise), specify that servers must never
append a "charset=" to a "text/html" content-type header, and add a
new tag to HTML which specifies the charset of the HTML. Then a server
that wants to rewrite charsets can implement whatever HTML parsing and
generation code that it happens to need, and one that doesn't can ignore
the issue completely.

I'm apparently not the only one that thinks this way, since this is
exactly how the xml content-types work.

cjs
-- 
Curt Sampson       <cjs@example.com>        +81 90 7737 2974   
Mobile sites and software consulting: http://www.starling-software.com