Last night I realised that if anybody went directly to
http://www.linux.ie/missing.html the
page would display - telling the user that it was missing! (For a comparison, open up the URL http://www.linux.ie/adskfjadsfkj/
(Webmaster's Note: This is a fine point,
likely to be lost on non-propeller heads. Ken is saying that since you're going
directly to the missing file error page, it should tell you that it _is_ there,
not that it's missing.)
So I am now using the following code to catch this, and handle it 'properly'.
<!--#if expr="\"$DOCUMENT_URI\" = \"$REQUEST_URI\"" -->
<TITLE>
Trying to be smart, are you?
</TITLE>
<!--#else -->
<TITLE>Missing file,
http://<!--#echo var="SERVER_NAME" --><!--#echo var="REQUEST_URI" -->
</TITLE>
<!--#endif -->
This works because (as you have documented on your site) DOCUMENT_URI contains the name of the page being served and REQUEST_URI contains the name of the document that was requested. So if the values of both are identical then the script [in the page] knows that the page was asked for instead of the page being served to the browser because some other one does not exist. Thus solving the existential crisis!
Webmaster's Note: Ken uses a couple of sets of the code above to display the correct message on the page.
