homepage Welcome to WebmasterWorld Guest from 184.72.82.126
register, free tools, login, search, pro membership, help, library, announcements, recent posts, open posts,
Become a Pro Member
Home / Forums Index / WebmasterWorld / Website Analytics - Tracking and Logging
Forum Library, Charter, Moderators: Receptional & mademetop

Website Analytics - Tracking and Logging Forum

    
GET /folder/null
Mozilla issue, not just GET /a_life
stevenha




msg:892542
 11:11 pm on Sep 27, 2003 (gmt 0)

This month, my access logs are showing more entries asking for GET /folder/null
I did a search on this forum, and found an old humorus thread on this, saying it was just some folks who needed to GET /a_life

But I'm not buying it. I had over 500 of these requests this month, all coming from various Mozilla UAs, and nearly all were referred by existing pages on my own site.

So can someone tell me, what is the cause of this? Is there a secret button in the browser that tries to show the contents of the current directory?

Should I be putting a rule in .htaccess to either help these users, or not?

 

moltar




msg:892543
 11:37 pm on Sep 27, 2003 (gmt 0)

Ya, I had lots of those too and they all, obviosly, caused 404s, that's how I found them.

stevenha




msg:892544
 2:11 am on Oct 2, 2003 (gmt 0)

I'm getting lots of these requests for /folder/null
and I still don't know why. I also ended up in a big mess when I tried to write a .htaccess rule to deal with them. Thus, I'm bumping this thread up to discuss again, in hopes that someone can help me explain and understand these requests, and suggest the right way to deal with them.

Yidaki




msg:892545
 10:06 am on Oct 24, 2003 (gmt 0)

Fwiw: i just found a bunch of /folder/nul/ GET requests in my error logs too.

These are the user agents:

- Mozilla/4.06 [en]C-gatewaynet (Win98; I)
- Mozilla/4.7 [de] (WinNT; U)
- Mozilla/4.51 [en] (X11; U; SunOS 5.8 sun4u)

The GET /a_life argument doesn't count imo. All requests are mixed with valid requests from the same ip's/ua's. Nothing suspicious.

Anbody know what's that?

stevenha




msg:892546
 5:13 am on Nov 1, 2003 (gmt 0)

Last month I had 1500 of these requests for */null files, and this month, there were 2000 of these file-not-found 404 errors in my logs.

Yidaki, I agree with you, that these seem to be coming from legitimate surfers, using Mozilla, on a variety of OS'es. I personally don't use Mozilla, so I'm again asking for help with this.

Can someone explain what feature or event occurs in Mozilla, that generates requests for GET /folder/null?

Could someone suggest a rewrite rule to automatically route these to the homepage? ( I'm a rookie at that stuff. is "null" a protected word in .htaccess files, or can a simple rule doe the trick?)

MonkeeSage




msg:892547
 5:37 am on Nov 1, 2003 (gmt 0)

Sounds to me like a script somewhere is doing something like this...

var varname = null;
// something that is supposed to set varname here, but fails
window.location.href = "/folder/" + varname;

Might be a bot or a script error on your own pages or something like that. Best guess I can come up with.

Jordan

stevenha




msg:892548
 6:04 am on Nov 1, 2003 (gmt 0)

Jordan, That's an interesting thought. I think you are definitely on the right track.

My logs show the referring URL is coming from a wide range of pages on my site. But, apart from Adsense, the vast majority of my pages don't use javascript.

Does Mozilla do something to automatically to invoke some scripts to set its BACK button destination, or other navigation buttons or features?

Anyone have any suggestions for rewrite rules to take care of these things?

MonkeeSage




msg:892549
 9:33 am on Nov 1, 2003 (gmt 0)

stevenha:

It's a good thought, because the Mozilla UI is all driven by JavaScript; but it is set up in such a way that errors in it are caught and dumped to the JS console -- they shouldn't ever be causing a GET request like that. I've been using K-Meleon / Mozilla / Firebird / Galeon for some time now and I usually run nightlies or my own builds off CVS (which are more prone to bugs), and I've never seen Gecko do anything like that. The worst I've seen is a crash, the most often is just an error message on the console -- never seen it send a request with a null in it.

I don't know much at all about mod_rewrite, so I can't be of any help there, but as a temporary solution you could make a blank file called 'null' in '/folder/' just to stop the 404s.

Jordan

bird




msg:892550
 12:27 pm on Nov 1, 2003 (gmt 0)

Looks like a broken proxy to me.

The requests are often mixed with seemingly "normal" request with the UA "Mozilla/3.01 (compatible;)", which is commonly used by some proxy software.

At first, I suspected a broken robot, but then I found a request to the search feature on my site following up to one of the broken requests. My 404 page prominently offers a search form, which uses a POST action, and is therefore very unlikely to be triggered by a robot. So everything points to a frustrated human trying to find the page he really wanted to go to in the first place.

Btw: The browser with the JavaScript driven UI is the new Mozilla. The UAs given are all in the NS 4.x range, which (I think) had its UI coded in plain C.

stevenha




msg:892551
 7:37 pm on Nov 1, 2003 (gmt 0)

I've been thinking about this, and I did a little more detective work in my logs. I've seen that users generally arrive on my normal static html pages, via normal mechanisms, like from search engines or from external links. I see them view the page and load the page's gifs. I see them click a link to a different page, which loads its its own gifs. Then, sometimes the succeed in going back to the previous page, but sometimes they start leaving log traces of these requests for GET /foldername/null and that suggests to me that they are probably trying to nagivate BACK, but failing.

They are ending up my my custom 404 page. I supposed I could (with help), modify .htaccess to route them someplace else... but that would confuse them too.. since they really wanted to go back.

Maybe I need to address this problem on my 404 page, with javascript, as in detect Mozilla, detect /null, and if so, do something clever. I'm afraid to be too clever though, because if their browsers javascript is screwing up already, further screwups may ensue...

I just wish I knew for sure, which button the users were pushing, that was causing this problem.

Yidaki




msg:892552
 11:32 am on Nov 2, 2003 (gmt 0)

>I just wish I knew for sure, which button the users were pushing, that was causing this problem.

I gonna set up a /null response page at one of my sites and ask the user to give me some feedback of why they get there - a quick form just with a message input box. I'm sure i'll get at least one constructive feedback. If so, i'll report back here.

stevenha




msg:892553
 2:00 am on Nov 22, 2003 (gmt 0)

I put up a special page to catch requests for folder/null, but after 173 pageviews, noone has responded to my request to receive an email from them, to tell me how they got there. But my logs show that user agents are various versions of Mozilla/4.0 up to Mozilla/4.79.

Has anyone else had any luck in figuring this out?

Global Options:
 top home search open messages active posts  
 

Home / Forums Index / WebmasterWorld / Website Analytics - Tracking and Logging
rss feed

All trademarks and copyrights held by respective owners. Member comments are owned by the poster.
Home ¦ Free Tools ¦ Terms of Service ¦ Privacy Policy ¦ Report Problem ¦ About ¦ Library ¦ Newsletter
WebmasterWorld is a Developer Shed Community owned by Jim Boykin.
© Webmaster World 1996-2014 all rights reserved