Forum Moderators: open
-G
I don't like the feature so I turned it off!
MSR
If you are running apache you could circumvent this "feature" by having all your 404's rerouted to a nph script which will print out a good 200 header and then print up what you want. Create your own backdoor instead of having it lead to ms. It is your traffic.
The nph-404 script could look something like this:
#!/usr/bin/perl
my $dt = gmtime(time);
$dt =~ s/ / 0/g ;
my $dt2 .= $dt;
my ($day,$month,$mday,$time,$year,) = split(/ /,$dt2);
print "$ENV{'SERVER_PROTOCOL'} 200 OK\r\nDate: $day\, $mday $month
$year $time GMT\r\nServer: $ENV{'SERVER_SOFTWARE'}\r\nConnecton: close\r\n";
print "Content-Type: text/html\n\n";print qq¦
<h1>Bill's not taking my money!!</h1><BR>
or whatever...
¦;
exit;
A thousand variations are possible, but that's the idea.
Of course the script won't do any good if IE is unable to resolve the domain name. That script is designed to circumvent IE's 404 splash.-- what it will display when a page isn't found.
When I first through that together I thought IE6 took all 404s over to that search box, not just unresolvable domains.
I loved the example used by cnet for a mistyped url..
microsoft.con!
[google.com...]
I made a change in the search feature of the browser so it defaults to Google. But wasn't too easy, a lot would probably let it go rather than bother.
Littleman's post is actually a reply to a question I asked on another thread, about "404 pages in IE5 and IE6," asking about what the script is for.
[webmasterworld.com...]
>>That script is designed to circumvent IE's 404 splash.-- what it will display when a page isn't found.<<
If I understand your reply correctly, does this mean that the script really isn't necessary if IE6 can resolve your domain? I'm trying to figure out what to tell clients to do about their 404s to cover recent versions of IE.
To avoid the "friendly" message in IE5, as I understand it, if you make your 404 at least 512-bytes, you're supposed to be OK. About IE6... on the thread, mdharrold reports:
I just tried it with a 189 byte file. IE 6 displayed my page.
Does this mean that a 512-byte 404 will cover it for both browsers, and there's nothing further a client should really be worrying about?
I am just waiting for the massively helpful targeted banner adds, and pop-ups using the lost name.
And since you don't need to go to that webmasterworld.com site for SE information, will just generate a 404 for you and provide all the appropriate information pointing to bCentral. What the heck, since we know what is best for you, we will let you type in any URL, but will use it as the search words to direct you to the right Microsoft site you must have wanted.
There are 404 errors that are generated when a document at a known domain cannot be found. The functionality of the MS browser can be overridden by commands used in .htaccess files, as explained above.
Then there are the DNS errors, where you are trying to access a domain that doesn't exist, for example by having a typo in the domain you're trying to access. This cannot be overridden by commands in .htaccess, because the surfer never accessed your domain in the first place.
Again, I'm not sure what the script is intended to do. For a document at a known domain, is the script necessary, or will a larger than 512-byte 404 suffice?
The script seems to be intended to thwart functionality of the MS browser that maybe doesn't exist... which is the appropriation of all 404s. As I'm now understanding it, the browser only displays the MSN search box when an unknown domain is requested.
Please forgive repeated asking of an apparently very simple question...
The script is not necessary for a known domain. But, if you want your less than 512-byte 404 page to be seen by older IE versions, the script will make that happen. It also gives a no-parsed-headers (nph) strater script for those without mod-rewrite.
mdharrold - Thanks... The script then seems like an elegant solution for a problem that doesn't really exist. ;)
