I'd like to know what content-type the server is sending. In
particular what character set is being used.
Does the server send the content-type by looking at the page extension
(mime types?) Or does it try to read file. The server in question is
If I use view info in NS7.1 it reports a character encoding. I'm
thinking that this is probably read from the document, not the char set
sent in the header, but I don't know. I'm unable to find page properties
in Opera and I'm unsure of what IE reports/
That'll get you the HTTP headers, but it might not get you a charset.
And even if it does somehow get you a charset, there's no guarantee
that it's not being overridden by the document, or by the client
software doing some heuristic nonsense.
Patron Saint of Drunken Fornication
It will get the charset if and only if the charset is specified in the
HTTP headers (in the Content-Type header).
> And even if it does somehow get you a charset, there's no guarantee
> that it's not being overridden by the document, or by the client
> software doing some heuristic nonsense.
There's a recent discussion in comp.infosystems.www.authoring.html
that has confirmed the observation that no browser is known to violate
the rule that charset specified in HTTP headers takes precedence over any
attempts to specify the charset inside the document itself (e.g., in
a <meta> tag).
If the headers don't specify the charset, browsers are allowed to (and
generally will) use charset specified in a <meta> tag or in some other
way. It gets fairly complex then, and at some point a browser is allowed
to (and generally will) make some wild guess ("use heuristics"),
typically letting the user try different encodings if the presentation
does not look right (but this may well exceed the user's capabilities).