Gone dotty

[Toby A Inkster]

[http://porjes.haxorz.org/one.x.y.z...html]

It appears that the server is not sending any MIME header at all.
Perhaps it has issues with determining the file type of the multi-dot
extension. That would confuse some browsers.

No, the server is sending a MIME type (text/html), but it is also sending
a Content-Encoding of 'x-compress', which is patently untrue. This is
because of the '.z' extension (although Apache can be configured to ignore
such silliness).

The full HTTP header:

[tai@ophelia (pts/1) ~]$ lynx -dump -head http://porjes.haxorz.org/one.x.y.z...html
HTTP/1.1 200 OK
Date: Sun, 03 Aug 2003 23:57:40 GMT
Server: Apache
Accept-Ranges: bytes
Content-Length: 190
Connection: close
Content-Type: text/html; charset=ISO-8859-1
Content-Encoding: x-compress

You see, Lynx is behaving correctly -- the server claims that the file is
compressed, so Lynx attempts to decompress it.

 

[Art Sackett]

The server uses the file extension to determine the MIME type sent
to the client.

In some cases, yes. In others, the server might look at the first few
bytes of the file to discover its type, in others it might just serve
everything in a given directory with the same Content-type: header.
But this would not account for the original problem, nor would it serve
to illustrate Tina's point, which I believe to be, um, mistaken.

The browser uses the MIME type to decide how to
handle the file eg. reassemble a gif or jpg, interpret the html,
hand a .doc file off to another application, etc.

This is not universally true. Some browsers have been known to
completely ignore Content-type: and rely wholly upon the file name
extension.

If the server can't figure out the file type, it most likely will
send the html MIME ... or perhaps just leave it up to the
browser to figure it out.

I've yet to see a server behave that way. Most will send along, if
nothing else, the default type, which is most often text/plain.

 

[Art Sackett]

Sure it's not your newsreader breaking the url?

I'm absolutely certain of it. I manually entered the URL into my
browser... and got it wrong. Just now when I entered it correctly, I
got the document entitled "helo".

Oops.